diff --git a/go.mod b/go.mod
index 6087e52d3e..614f3639f8 100644
--- a/go.mod
+++ b/go.mod
@@ -140,4 +140,4 @@ require (
 	xorm.io/core v0.6.3
 )
 
-replace github.com/denisenkom/go-mssqldb => github.com/denisenkom/go-mssqldb v0.0.0-20161128230840-e32ca5036449
+replace github.com/denisenkom/go-mssqldb => github.com/denisenkom/go-mssqldb v0.0.0-20180314172330-6a30f4e59a44
diff --git a/go.sum b/go.sum
index 9d5e2c9c50..f34fe7f3ec 100644
--- a/go.sum
+++ b/go.sum
@@ -60,8 +60,8 @@ github.com/cznic/strutil v0.0.0-20181122101858-275e90344537/go.mod h1:AHHPPPXTw0
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
-github.com/denisenkom/go-mssqldb v0.0.0-20161128230840-e32ca5036449 h1:JpA+YMG4JLW8nzLmU05mTiuB0O17xHGxpWolEZ0zDuA=
-github.com/denisenkom/go-mssqldb v0.0.0-20161128230840-e32ca5036449/go.mod h1:xN/JuLBIz4bjkxNmByTiV1IbhfnYb6oo99phBn4Eqhc=
+github.com/denisenkom/go-mssqldb v0.0.0-20180314172330-6a30f4e59a44 h1:x0uHqLQTSEL9LKic8sWDt3ASkq07ve5ojIIUl5uF64M=
+github.com/denisenkom/go-mssqldb v0.0.0-20180314172330-6a30f4e59a44/go.mod h1:xN/JuLBIz4bjkxNmByTiV1IbhfnYb6oo99phBn4Eqhc=
 github.com/dgrijalva/jwt-go v0.0.0-20161101193935-9ed569b5d1ac h1:xrQJVwQCGqDvOO7/0+RyIq5J2M3Q4ZF7Ug/BMQtML1E=
 github.com/dgrijalva/jwt-go v0.0.0-20161101193935-9ed569b5d1ac/go.mod h1:E3ru+11k8xSBh+hMPgOLZmtrrCbhqsmaPHjLKYnJCaQ=
 github.com/edsrzf/mmap-go v0.0.0-20170320065105-0bce6a688712 h1:aaQcKT9WumO6JEJcRyTqFVq4XUZiUcKR2/GI31TOcz8=
diff --git a/vendor/github.com/denisenkom/go-mssqldb/README.md b/vendor/github.com/denisenkom/go-mssqldb/README.md
index 8570ae9f61..cd19ded516 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/README.md
+++ b/vendor/github.com/denisenkom/go-mssqldb/README.md
@@ -1,79 +1,131 @@
 # A pure Go MSSQL driver for Go's database/sql package
 
+[![GoDoc](https://godoc.org/github.com/denisenkom/go-mssqldb?status.svg)](http://godoc.org/github.com/denisenkom/go-mssqldb)
+[![Build status](https://ci.appveyor.com/api/projects/status/jrln8cs62wj9i0a2?svg=true)](https://ci.appveyor.com/project/denisenkom/go-mssqldb)
+[![codecov](https://codecov.io/gh/denisenkom/go-mssqldb/branch/master/graph/badge.svg)](https://codecov.io/gh/denisenkom/go-mssqldb)
+
 ## Install
 
-    go get github.com/denisenkom/go-mssqldb
+Requires Go 1.8 or above.
 
-## Tests
+Install with `go get github.com/denisenkom/go-mssqldb` .
 
-`go test` is used for testing. A running instance of MSSQL server is required.
-Environment variables are used to pass login information.
+## Connection Parameters and DSN
 
-Example:
+The recommended connection string uses a URL format:
+`sqlserver://username:password@host/instance?param1=value&param2=value`
+Other supported formats are listed below.
 
-    env HOST=localhost SQLUSER=sa SQLPASSWORD=sa DATABASE=test go test
+### Common parameters:
 
-## Connection Parameters
+* `user id` - enter the SQL Server Authentication user id or the Windows Authentication user id in the DOMAIN\User format. On Windows, if user id is empty or missing Single-Sign-On is used.
+* `password`
+* `database`
+* `connection timeout` - in seconds (default is 30)
+* `dial timeout` - in seconds (default is 5)
+* `encrypt`
+  * `disable` - Data send between client and server is not encrypted.
+  * `false` - Data sent between client and server is not encrypted beyond the login packet. (Default)
+  * `true` - Data sent between client and server is encrypted.
+* `keepAlive` - in seconds; 0 to disable (default is 30)
+* `app name` - The application name (default is go-mssqldb)
 
-* "server" - host or host\instance (default localhost)
-* "port" - used only when there is no instance in server (default 1433)
-* "failoverpartner" - host or host\instance (default is no partner). 
-* "failoverport" - used only when there is no instance in failoverpartner (default 1433)
-* "user id" - enter the SQL Server Authentication user id or the Windows Authentication user id in the DOMAIN\User format. On Windows, if user id is empty or missing Single-Sign-On is used.
-* "password"
-* "database"
-* "connection timeout" - in seconds (default is 30)
-* "dial timeout" - in seconds (default is 5)
-* "keepAlive" - in seconds; 0 to disable (default is 0)
-* "log" - logging flags (default 0/no logging, 63 for full logging)
+### Connection parameters for ODBC and ADO style connection strings:
+
+* `server` - host or host\instance (default localhost)
+* `port` - used only when there is no instance in server (default 1433)
+
+### Less common parameters:
+
+* `failoverpartner` - host or host\instance (default is no partner). 
+* `failoverport` - used only when there is no instance in failoverpartner (default 1433)
+* `packet size` - in bytes; 512 to 32767 (default is 4096)
+  * Encrypted connections have a maximum packet size of 16383 bytes
+  * Further information on usage: https://docs.microsoft.com/en-us/sql/database-engine/configure-windows/configure-the-network-packet-size-server-configuration-option
+* `log` - logging flags (default 0/no logging, 63 for full logging)
   *  1 log errors
   *  2 log messages
   *  4 log rows affected
   *  8 trace sql statements
   * 16 log statement parameters
   * 32 log transaction begin/end
-* "encrypt"
-  * disable - Data send between client and server is not encrypted.
-  * false - Data sent between client and server is not encrypted beyond the login packet. (Default)
-  * true - Data sent between client and server is encrypted.
-* "TrustServerCertificate"
+* `TrustServerCertificate`
   * false - Server certificate is checked. Default is false if encypt is specified.
   * true - Server certificate is not checked. Default is true if encrypt is not specified. If trust server certificate is true, driver accepts any certificate presented by the server and any host name in that certificate. In this mode, TLS is susceptible to man-in-the-middle attacks. This should be used only for testing.
-* "certificate" - The file that contains the public key certificate of the CA that signed the SQL Server certificate. The specified certificate overrides the go platform specific CA certificates.
-* "hostNameInCertificate" - Specifies the Common Name (CN) in the server certificate. Default value is the server host.
-* "ServerSPN" - The kerberos SPN (Service Principal Name) for the server. Default is MSSQLSvc/host:port.
-* "Workstation ID" - The workstation name (default is the host name)
-* "app name" - The application name (default is go-mssqldb)
-* "ApplicationIntent" - Can be given the value "ReadOnly" to initiate a read-only connection to an Availability Group listener.
+* `certificate` - The file that contains the public key certificate of the CA that signed the SQL Server certificate. The specified certificate overrides the go platform specific CA certificates.
+* `hostNameInCertificate` - Specifies the Common Name (CN) in the server certificate. Default value is the server host.
+* `ServerSPN` - The kerberos SPN (Service Principal Name) for the server. Default is MSSQLSvc/host:port.
+* `Workstation ID` - The workstation name (default is the host name)
+* `ApplicationIntent` - Can be given the value `ReadOnly` to initiate a read-only connection to an Availability Group listener.
 
-Example:
+### The connection string can be specified in one of three formats:
+
+
+1. URL: with `sqlserver` scheme. username and password appears before the host. Any instance appears as
+    the first segment in the path. All other options are query parameters. Examples:
+
+  * `sqlserver://username:password@host/instance?param1=value&param2=value`
+  * `sqlserver://username:password@host:port?param1=value&param2=value`
+  * `sqlserver://sa@localhost/SQLExpress?database=master&connection+timeout=30` // `SQLExpress instance.
+  * `sqlserver://sa:mypass@localhost?database=master&connection+timeout=30`     // username=sa, password=mypass.
+  * `sqlserver://sa:mypass@localhost:1234?database=master&connection+timeout=30"` // port 1234 on localhost.
+  * `sqlserver://sa:my%7Bpass@somehost?connection+timeout=30` // password is "my{pass"
+
+  A string of this format can be constructed using the `URL` type in the `net/url` package.
 
 ```go
-    db, err := sql.Open("mssql", "server=localhost;user id=sa")
+  query := url.Values{}
+  query.Add("connection timeout", "30")
+
+  u := &url.URL{
+      Scheme:   "sqlserver",
+      User:     url.UserPassword(username, password),
+      Host:     fmt.Sprintf("%s:%d", hostname, port),
+      // Path:  instance, // if connecting to an instance instead of a port
+      RawQuery: query.Encode(),
+  }
+  db, err := sql.Open("sqlserver", u.String())
+```
+
+2. ADO: `key=value` pairs separated by `;`. Values may not contain `;`, leading and trailing whitespace is ignored.
+     Examples:
+	
+  * `server=localhost\\SQLExpress;user id=sa;database=master;connection timeout=30`
+  * `server=localhost;user id=sa;database=master;connection timeout=30`
+
+3. ODBC: Prefix with `odbc`, `key=value` pairs separated by `;`. Allow `;` by wrapping
+    values in `{}`. Examples:
+	
+  * `odbc:server=localhost\\SQLExpress;user id=sa;database=master;connection timeout=30`
+  * `odbc:server=localhost;user id=sa;database=master;connection timeout=30`
+  * `odbc:server=localhost;user id=sa;password={foo;bar}` // Value marked with `{}`, password is "foo;bar"
+  * `odbc:server=localhost;user id=sa;password={foo{bar}` // Value marked with `{}`, password is "foo{bar"
+  * `odbc:server=localhost;user id=sa;password={foobar }` // Value marked with `{}`, password is "foobar "
+  * `odbc:server=localhost;user id=sa;password=foo{bar`   // Literal `{`, password is "foo{bar"
+  * `odbc:server=localhost;user id=sa;password=foo}bar`   // Literal `}`, password is "foo}bar"
+  * `odbc:server=localhost;user id=sa;password={foo{bar}` // Literal `{`, password is "foo{bar"
+  * `odbc:server=localhost;user id=sa;password={foo}}bar}` // Escaped `} with `}}`, password is "foo}bar"
+
+## Executing Stored Procedures
+
+To run a stored procedure, set the query text to the procedure name:
+```go
+var account = "abc"
+_, err := db.ExecContext(ctx, "sp_RunMe",
+	sql.Named("ID", 123),
+	sql.Out{Dest{sql.Named("Account", &account)}
+)
 ```
 
 ## Statement Parameters
 
-In the SQL statement text, literals may be replaced by a parameter that matches one of the following:
-
-* ?
-* ?nnn
-* :nnn
-* $nnn
-
-where nnn represents an integer that specifies a 1-indexed positional parameter. Ex:
+The `sqlserver` driver uses normal MS SQL Server syntax and expects parameters in
+the sql query to be in the form of either `@Name` or `@p1` to `@pN` (ordinal position).
 
 ```go
-db.Query("SELECT * FROM t WHERE a = ?3, b = ?2, c = ?1", "x", "y", "z")
+db.QueryContext(ctx, `select * from t where ID = @ID and Name = @p2;`, sql.Named("ID", 6), "Bob")
 ```
 
-will expand to roughly
-
-```sql
-SELECT * FROM t WHERE a = 'z', b = 'y', c = 'x'
-```
-
-
 ## Features
 
 * Can be used with SQL Server 2005 or newer
@@ -87,6 +139,34 @@ SELECT * FROM t WHERE a = 'z', b = 'y', c = 'x'
 * Supports connections to AlwaysOn Availability Group listeners, including re-direction to read-only replicas.
 * Supports query notifications
 
+## Tests
+
+`go test` is used for testing. A running instance of MSSQL server is required.
+Environment variables are used to pass login information.
+
+Example:
+
+    env SQLSERVER_DSN=sqlserver://user:pass@hostname/instance?database=test1 go test
+
+## Deprecated
+
+These features still exist in the driver, but they are are deprecated.
+
+### Query Parameter Token Replace (driver "mssql")
+
+If you use the driver name "mssql" (rather then "sqlserver" the SQL text
+will be loosly parsed and an attempt to extract identifiers using one of
+
+* ?
+* ?nnn
+* :nnn
+* $nnn
+
+will be used. This is not recommended with SQL Server. 
+There is at least one existing `won't fix` issue with the query parsing.
+
+Use the native "@Name" parameters instead with the "sqlserver" driver name.
+
 ## Known Issues
 
 * SQL Server 2008 and 2008 R2 engine cannot handle login records when SSL encryption is not disabled.
diff --git a/vendor/github.com/denisenkom/go-mssqldb/appveyor.yml b/vendor/github.com/denisenkom/go-mssqldb/appveyor.yml
new file mode 100644
index 0000000000..2c1435507b
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/appveyor.yml
@@ -0,0 +1,45 @@
+version: 1.0.{build}
+
+os: Windows Server 2012 R2
+
+clone_folder: c:\gopath\src\github.com\denisenkom\go-mssqldb
+
+environment:
+  GOPATH: c:\gopath
+  HOST: localhost
+  SQLUSER: sa
+  SQLPASSWORD: Password12!
+  DATABASE: test  
+  GOVERSION: 110
+  matrix:
+    - GOVERSION: 18
+      SQLINSTANCE: SQL2016
+    - GOVERSION: 110
+      SQLINSTANCE: SQL2016
+    - SQLINSTANCE: SQL2014
+    - SQLINSTANCE: SQL2012SP1
+    - SQLINSTANCE: SQL2008R2SP2
+ 
+install:
+  - set GOROOT=c:\go%GOVERSION%
+  - set PATH=%GOPATH%\bin;%GOROOT%\bin;%PATH%
+  - go version
+  - go env
+
+build_script:
+  - go build
+
+before_test:
+  # setup SQL Server 
+  - ps: | 
+      $instanceName = $env:SQLINSTANCE
+      Start-Service "MSSQL`$$instanceName"
+      Start-Service "SQLBrowser"
+  - sqlcmd -S "(local)\%SQLINSTANCE%" -Q "Use [master]; CREATE DATABASE test;"
+  - sqlcmd -S "(local)\%SQLINSTANCE%" -h -1 -Q "set nocount on; Select @@version"
+  - pip install codecov
+ 
+
+test_script:
+  - go test -race -coverprofile=coverage.txt -covermode=atomic
+  - codecov -f coverage.txt
diff --git a/vendor/github.com/denisenkom/go-mssqldb/buf.go b/vendor/github.com/denisenkom/go-mssqldb/buf.go
index 42e8ae345c..365acd4833 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/buf.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/buf.go
@@ -2,12 +2,14 @@ package mssql
 
 import (
 	"encoding/binary"
-	"io"
 	"errors"
+	"io"
 )
 
+type packetType uint8
+
 type header struct {
-	PacketType uint8
+	PacketType packetType
 	Status     uint8
 	Size       uint16
 	Spid       uint16
@@ -15,55 +17,84 @@ type header struct {
 	Pad        uint8
 }
 
+// tdsBuffer reads and writes TDS packets of data to the transport.
+// The write and read buffers are separate to make sending attn signals
+// possible without locks. Currently attn signals are only sent during
+// reads, not writes.
 type tdsBuffer struct {
-	buf         []byte
-	pos         uint16
-	transport   io.ReadWriteCloser
-	size        uint16
+	transport io.ReadWriteCloser
+
+	packetSize int
+
+	// Write fields.
+	wbuf        []byte
+	wpos        int
+	wPacketSeq  byte
+	wPacketType packetType
+
+	// Read fields.
+	rbuf        []byte
+	rpos        int
+	rsize       int
 	final       bool
-	packet_type uint8
-	afterFirst  func()
+	rPacketType packetType
+
+	// afterFirst is assigned to right after tdsBuffer is created and
+	// before the first use. It is executed after the first packet is
+	// written and then removed.
+	afterFirst func()
 }
 
-func newTdsBuffer(bufsize int, transport io.ReadWriteCloser) *tdsBuffer {
-	buf := make([]byte, bufsize)
-	w := new(tdsBuffer)
-	w.buf = buf
-	w.pos = 8
-	w.transport = transport
-	w.size = 0
-	return w
+func newTdsBuffer(bufsize uint16, transport io.ReadWriteCloser) *tdsBuffer {
+	return &tdsBuffer{
+		packetSize: int(bufsize),
+		wbuf:       make([]byte, 1<<16),
+		rbuf:       make([]byte, 1<<16),
+		rpos:       8,
+		transport:  transport,
+	}
+}
+
+func (rw *tdsBuffer) ResizeBuffer(packetSize int) {
+	rw.packetSize = packetSize
+}
+
+func (w *tdsBuffer) PackageSize() int {
+	return w.packetSize
 }
 
 func (w *tdsBuffer) flush() (err error) {
-	// writing packet size
-	binary.BigEndian.PutUint16(w.buf[2:], w.pos)
+	// Write packet size.
+	w.wbuf[0] = byte(w.wPacketType)
+	binary.BigEndian.PutUint16(w.wbuf[2:], uint16(w.wpos))
+	w.wbuf[6] = w.wPacketSeq
 
-	// writing packet into underlying transport
-	if _, err = w.transport.Write(w.buf[:w.pos]); err != nil {
+	// Write packet into underlying transport.
+	if _, err = w.transport.Write(w.wbuf[:w.wpos]); err != nil {
 		return err
 	}
+	// It is possible to create a whole new buffer after a flush.
+	// Useful for debugging. Normally reuse the buffer.
+	// w.wbuf = make([]byte, 1<<16)
 
-	// execute afterFirst hook if it is set
+	// Execute afterFirst hook if it is set.
 	if w.afterFirst != nil {
 		w.afterFirst()
 		w.afterFirst = nil
 	}
 
-	w.pos = 8
-	// packet number
-	w.buf[6] += 1
+	w.wpos = 8
+	w.wPacketSeq++
 	return nil
 }
 
 func (w *tdsBuffer) Write(p []byte) (total int, err error) {
-	total = 0
 	for {
-		copied := copy(w.buf[w.pos:], p)
-		w.pos += uint16(copied)
+		copied := copy(w.wbuf[w.wpos:w.packetSize], p)
+		w.wpos += copied
 		total += copied
 		if copied == len(p) {
-			break
+			return
 		}
 		if err = w.flush(); err != nil {
 			return
@@ -74,66 +105,64 @@ func (w *tdsBuffer) Write(p []byte) (total int, err error) {
 }
 
 func (w *tdsBuffer) WriteByte(b byte) error {
-	if int(w.pos) == len(w.buf) {
+	if int(w.wpos) == len(w.wbuf) {
 		if err := w.flush(); err != nil {
 			return err
 		}
 	}
-	w.buf[w.pos] = b
-	w.pos += 1
+	w.wbuf[w.wpos] = b
+	w.wpos += 1
 	return nil
 }
 
-func (w *tdsBuffer) BeginPacket(packet_type byte) {
-	w.buf[0] = packet_type
-	w.buf[1] = 0 // packet is incomplete
-	w.buf[4] = 0 // spid
-	w.buf[5] = 0
-	w.buf[6] = 1 // packet id
-	w.buf[7] = 0 // window
-	w.pos = 8
+func (w *tdsBuffer) BeginPacket(packetType packetType) {
+	w.wbuf[1] = 0 // Packet is incomplete. This byte is set again in FinishPacket.
+	w.wpos = 8
+	w.wPacketSeq = 1
+	w.wPacketType = packetType
 }
 
 func (w *tdsBuffer) FinishPacket() error {
-	w.buf[1] = 1 // this is last packet
+	w.wbuf[1] = 1 // Mark this as the last packet in the message.
 	return w.flush()
 }
 
+var headerSize = binary.Size(header{})
+
 func (r *tdsBuffer) readNextPacket() error {
-	header := header{}
+	h := header{}
 	var err error
-	err = binary.Read(r.transport, binary.BigEndian, &header)
+	err = binary.Read(r.transport, binary.BigEndian, &h)
 	if err != nil {
 		return err
 	}
-	offset := uint16(binary.Size(header))
-	if int(header.Size) > len(r.buf) {
+	if int(h.Size) > len(r.rbuf) {
 		return errors.New("Invalid packet size, it is longer than buffer size")
 	}
-	if int(offset) > int(header.Size) {
+	if headerSize > int(h.Size) {
 		return errors.New("Invalid packet size, it is shorter than header size")
 	}
-	_, err = io.ReadFull(r.transport, r.buf[offset:header.Size])
+	_, err = io.ReadFull(r.transport, r.rbuf[headerSize:h.Size])
 	if err != nil {
 		return err
 	}
-	r.pos = offset
-	r.size = header.Size
-	r.final = header.Status != 0
-	r.packet_type = header.PacketType
+	r.rpos = headerSize
+	r.rsize = int(h.Size)
+	r.final = h.Status != 0
+	r.rPacketType = h.PacketType
 	return nil
 }
 
-func (r *tdsBuffer) BeginRead() (uint8, error) {
+func (r *tdsBuffer) BeginRead() (packetType, error) {
 	err := r.readNextPacket()
 	if err != nil {
 		return 0, err
 	}
-	return r.packet_type, nil
+	return r.rPacketType, nil
 }
 
 func (r *tdsBuffer) ReadByte() (res byte, err error) {
-	if r.pos == r.size {
+	if r.rpos == r.rsize {
 		if r.final {
 			return 0, io.EOF
 		}
@@ -142,8 +171,8 @@ func (r *tdsBuffer) ReadByte() (res byte, err error) {
 			return 0, err
 		}
 	}
-	res = r.buf[r.pos]
-	r.pos++
+	res = r.rbuf[r.rpos]
+	r.rpos++
 	return res, nil
 }
 
@@ -207,7 +236,7 @@ func (r *tdsBuffer) readUcs2(numchars int) string {
 func (r *tdsBuffer) Read(buf []byte) (copied int, err error) {
 	copied = 0
 	err = nil
-	if r.pos == r.size {
+	if r.rpos == r.rsize {
 		if r.final {
 			return 0, io.EOF
 		}
@@ -216,7 +245,7 @@ func (r *tdsBuffer) Read(buf []byte) (copied int, err error) {
 			return
 		}
 	}
-	copied = copy(buf, r.buf[r.pos:r.size])
-	r.pos += uint16(copied)
+	copied = copy(buf, r.rbuf[r.rpos:r.rsize])
+	r.rpos += copied
 	return
 }
diff --git a/vendor/github.com/denisenkom/go-mssqldb/bulkcopy.go b/vendor/github.com/denisenkom/go-mssqldb/bulkcopy.go
new file mode 100644
index 0000000000..8c0a4e0a2a
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/bulkcopy.go
@@ -0,0 +1,616 @@
+package mssql
+
+import (
+	"bytes"
+	"context"
+	"encoding/binary"
+	"fmt"
+	"math"
+	"reflect"
+	"strconv"
+	"strings"
+	"time"
+)
+
+type Bulk struct {
+	cn          *Conn
+	metadata    []columnStruct
+	bulkColumns []columnStruct
+	columnsName []string
+	tablename   string
+	numRows     int
+
+	headerSent bool
+	Options    BulkOptions
+	Debug      bool
+}
+type BulkOptions struct {
+	CheckConstraints  bool
+	FireTriggers      bool
+	KeepNulls         bool
+	KilobytesPerBatch int
+	RowsPerBatch      int
+	Order             []string
+	Tablock           bool
+}
+
+type DataValue interface{}
+
+func (cn *Conn) CreateBulk(table string, columns []string) (_ *Bulk) {
+	b := Bulk{cn: cn, tablename: table, headerSent: false, columnsName: columns}
+	b.Debug = false
+	return &b
+}
+
+func (b *Bulk) sendBulkCommand() (err error) {
+	//get table columns info
+	err = b.getMetadata()
+	if err != nil {
+		return err
+	}
+
+	//match the columns
+	for _, colname := range b.columnsName {
+		var bulkCol *columnStruct
+
+		for _, m := range b.metadata {
+			if m.ColName == colname {
+				bulkCol = &m
+				break
+			}
+		}
+		if bulkCol != nil {
+
+			if bulkCol.ti.TypeId == typeUdt {
+				//send udt as binary
+				bulkCol.ti.TypeId = typeBigVarBin
+			}
+			b.bulkColumns = append(b.bulkColumns, *bulkCol)
+			b.dlogf("Adding column %s %s %#x", colname, bulkCol.ColName, bulkCol.ti.TypeId)
+		} else {
+			return fmt.Errorf("Column %s does not exist in destination table %s", colname, b.tablename)
+		}
+	}
+
+	//create the bulk command
+
+	//columns definitions
+	var col_defs bytes.Buffer
+	for i, col := range b.bulkColumns {
+		if i != 0 {
+			col_defs.WriteString(", ")
+		}
+		col_defs.WriteString("[" + col.ColName + "] " + makeDecl(col.ti))
+	}
+
+	//options
+	var with_opts []string
+
+	if b.Options.CheckConstraints {
+		with_opts = append(with_opts, "CHECK_CONSTRAINTS")
+	}
+	if b.Options.FireTriggers {
+		with_opts = append(with_opts, "FIRE_TRIGGERS")
+	}
+	if b.Options.KeepNulls {
+		with_opts = append(with_opts, "KEEP_NULLS")
+	}
+	if b.Options.KilobytesPerBatch > 0 {
+		with_opts = append(with_opts, fmt.Sprintf("KILOBYTES_PER_BATCH = %d", b.Options.KilobytesPerBatch))
+	}
+	if b.Options.RowsPerBatch > 0 {
+		with_opts = append(with_opts, fmt.Sprintf("ROWS_PER_BATCH = %d", b.Options.RowsPerBatch))
+	}
+	if len(b.Options.Order) > 0 {
+		with_opts = append(with_opts, fmt.Sprintf("ORDER(%s)", strings.Join(b.Options.Order, ",")))
+	}
+	if b.Options.Tablock {
+		with_opts = append(with_opts, "TABLOCK")
+	}
+	var with_part string
+	if len(with_opts) > 0 {
+		with_part = fmt.Sprintf("WITH (%s)", strings.Join(with_opts, ","))
+	}
+
+	query := fmt.Sprintf("INSERT BULK %s (%s) %s", b.tablename, col_defs.String(), with_part)
+
+	stmt, err := b.cn.Prepare(query)
+	if err != nil {
+		return fmt.Errorf("Prepare failed: %s", err.Error())
+	}
+	b.dlogf(query)
+
+	_, err = stmt.Exec(nil)
+	if err != nil {
+		return err
+	}
+
+	b.headerSent = true
+
+	var buf = b.cn.sess.buf
+	buf.BeginPacket(packBulkLoadBCP)
+
+	// send the columns metadata
+	columnMetadata := b.createColMetadata()
+	_, err = buf.Write(columnMetadata)
+
+	return
+}
+
+// AddRow immediately writes the row to the destination table.
+// The arguments are the row values in the order they were specified.
+func (b *Bulk) AddRow(row []interface{}) (err error) {
+	if !b.headerSent {
+		err = b.sendBulkCommand()
+		if err != nil {
+			return
+		}
+	}
+
+	if len(row) != len(b.bulkColumns) {
+		return fmt.Errorf("Row does not have the same number of columns than the destination table %d %d",
+			len(row), len(b.bulkColumns))
+	}
+
+	bytes, err := b.makeRowData(row)
+	if err != nil {
+		return
+	}
+
+	_, err = b.cn.sess.buf.Write(bytes)
+	if err != nil {
+		return
+	}
+
+	b.numRows = b.numRows + 1
+	return
+}
+
+func (b *Bulk) makeRowData(row []interface{}) ([]byte, error) {
+	buf := new(bytes.Buffer)
+	buf.WriteByte(byte(tokenRow))
+
+	var logcol bytes.Buffer
+	for i, col := range b.bulkColumns {
+
+		if b.Debug {
+			logcol.WriteString(fmt.Sprintf(" col[%d]='%v' ", i, row[i]))
+		}
+		param, err := b.makeParam(row[i], col)
+		if err != nil {
+			return nil, fmt.Errorf("bulkcopy: %s", err.Error())
+		}
+
+		if col.ti.Writer == nil {
+			return nil, fmt.Errorf("no writer for column: %s, TypeId: %#x",
+				col.ColName, col.ti.TypeId)
+		}
+		err = col.ti.Writer(buf, param.ti, param.buffer)
+		if err != nil {
+			return nil, fmt.Errorf("bulkcopy: %s", err.Error())
+		}
+	}
+
+	b.dlogf("row[%d] %s\n", b.numRows, logcol.String())
+
+	return buf.Bytes(), nil
+}
+
+func (b *Bulk) Done() (rowcount int64, err error) {
+	if b.headerSent == false {
+		//no rows had been sent
+		return 0, nil
+	}
+	var buf = b.cn.sess.buf
+	buf.WriteByte(byte(tokenDone))
+
+	binary.Write(buf, binary.LittleEndian, uint16(doneFinal))
+	binary.Write(buf, binary.LittleEndian, uint16(0)) //     curcmd
+
+	if b.cn.sess.loginAck.TDSVersion >= verTDS72 {
+		binary.Write(buf, binary.LittleEndian, uint64(0)) //rowcount 0
+	} else {
+		binary.Write(buf, binary.LittleEndian, uint32(0)) //rowcount 0
+	}
+
+	buf.FinishPacket()
+
+	tokchan := make(chan tokenStruct, 5)
+	go processResponse(context.Background(), b.cn.sess, tokchan, nil)
+
+	var rowCount int64
+	for token := range tokchan {
+		switch token := token.(type) {
+		case doneStruct:
+			if token.Status&doneCount != 0 {
+				rowCount = int64(token.RowCount)
+			}
+			if token.isError() {
+				return 0, token.getError()
+			}
+		case error:
+			return 0, b.cn.checkBadConn(token)
+		}
+	}
+	return rowCount, nil
+}
+
+func (b *Bulk) createColMetadata() []byte {
+	buf := new(bytes.Buffer)
+	buf.WriteByte(byte(tokenColMetadata))                              // token
+	binary.Write(buf, binary.LittleEndian, uint16(len(b.bulkColumns))) // column count
+
+	for i, col := range b.bulkColumns {
+
+		if b.cn.sess.loginAck.TDSVersion >= verTDS72 {
+			binary.Write(buf, binary.LittleEndian, uint32(col.UserType)) //  usertype, always 0?
+		} else {
+			binary.Write(buf, binary.LittleEndian, uint16(col.UserType))
+		}
+		binary.Write(buf, binary.LittleEndian, uint16(col.Flags))
+
+		writeTypeInfo(buf, &b.bulkColumns[i].ti)
+
+		if col.ti.TypeId == typeNText ||
+			col.ti.TypeId == typeText ||
+			col.ti.TypeId == typeImage {
+
+			tablename_ucs2 := str2ucs2(b.tablename)
+			binary.Write(buf, binary.LittleEndian, uint16(len(tablename_ucs2)/2))
+			buf.Write(tablename_ucs2)
+		}
+		colname_ucs2 := str2ucs2(col.ColName)
+		buf.WriteByte(uint8(len(colname_ucs2) / 2))
+		buf.Write(colname_ucs2)
+	}
+
+	return buf.Bytes()
+}
+
+func (b *Bulk) getMetadata() (err error) {
+	stmt, err := b.cn.Prepare("SET FMTONLY ON")
+	if err != nil {
+		return
+	}
+
+	_, err = stmt.Exec(nil)
+	if err != nil {
+		return
+	}
+
+	//get columns info
+	stmt, err = b.cn.Prepare(fmt.Sprintf("select * from %s SET FMTONLY OFF", b.tablename))
+	if err != nil {
+		return
+	}
+	stmt2 := stmt.(*Stmt)
+	cols, err := stmt2.QueryMeta()
+	if err != nil {
+		return fmt.Errorf("get columns info failed: %v", err.Error())
+	}
+	b.metadata = cols
+
+	if b.Debug {
+		for _, col := range b.metadata {
+			b.dlogf("col: %s typeId: %#x size: %d scale: %d prec: %d flags: %d lcid: %#x\n",
+				col.ColName, col.ti.TypeId, col.ti.Size, col.ti.Scale, col.ti.Prec,
+				col.Flags, col.ti.Collation.LcidAndFlags)
+		}
+	}
+
+	return nil
+}
+
+// QueryMeta is almost the same as mssql.Stmt.Query, but returns all the columns info.
+func (s *Stmt) QueryMeta() (cols []columnStruct, err error) {
+	if err = s.sendQuery(nil); err != nil {
+		return
+	}
+	tokchan := make(chan tokenStruct, 5)
+	go processResponse(context.Background(), s.c.sess, tokchan, s.c.outs)
+	s.c.clearOuts()
+loop:
+	for tok := range tokchan {
+		switch token := tok.(type) {
+		case doneStruct:
+			break loop
+		case []columnStruct:
+			cols = token
+			break loop
+		case error:
+			return nil, s.c.checkBadConn(token)
+		}
+	}
+	return cols, nil
+}
+
+func (b *Bulk) makeParam(val DataValue, col columnStruct) (res Param, err error) {
+	res.ti.Size = col.ti.Size
+	res.ti.TypeId = col.ti.TypeId
+
+	if val == nil {
+		res.ti.Size = 0
+		return
+	}
+
+	switch col.ti.TypeId {
+
+	case typeInt1, typeInt2, typeInt4, typeInt8, typeIntN:
+		var intvalue int64
+
+		switch val := val.(type) {
+		case int:
+			intvalue = int64(val)
+		case int32:
+			intvalue = int64(val)
+		case int64:
+			intvalue = val
+		default:
+			err = fmt.Errorf("mssql: invalid type for int column")
+			return
+		}
+
+		res.buffer = make([]byte, res.ti.Size)
+		if col.ti.Size == 1 {
+			res.buffer[0] = byte(intvalue)
+		} else if col.ti.Size == 2 {
+			binary.LittleEndian.PutUint16(res.buffer, uint16(intvalue))
+		} else if col.ti.Size == 4 {
+			binary.LittleEndian.PutUint32(res.buffer, uint32(intvalue))
+		} else if col.ti.Size == 8 {
+			binary.LittleEndian.PutUint64(res.buffer, uint64(intvalue))
+		}
+	case typeFlt4, typeFlt8, typeFltN:
+		var floatvalue float64
+
+		switch val := val.(type) {
+		case float32:
+			floatvalue = float64(val)
+		case float64:
+			floatvalue = val
+		case int:
+			floatvalue = float64(val)
+		case int64:
+			floatvalue = float64(val)
+		default:
+			err = fmt.Errorf("mssql: invalid type for float column: %s", val)
+			return
+		}
+
+		if col.ti.Size == 4 {
+			res.buffer = make([]byte, 4)
+			binary.LittleEndian.PutUint32(res.buffer, math.Float32bits(float32(floatvalue)))
+		} else if col.ti.Size == 8 {
+			res.buffer = make([]byte, 8)
+			binary.LittleEndian.PutUint64(res.buffer, math.Float64bits(floatvalue))
+		}
+	case typeNVarChar, typeNText, typeNChar:
+
+		switch val := val.(type) {
+		case string:
+			res.buffer = str2ucs2(val)
+		case []byte:
+			res.buffer = val
+		default:
+			err = fmt.Errorf("mssql: invalid type for nvarchar column: %s", val)
+			return
+		}
+		res.ti.Size = len(res.buffer)
+
+	case typeVarChar, typeBigVarChar, typeText, typeChar, typeBigChar:
+		switch val := val.(type) {
+		case string:
+			res.buffer = []byte(val)
+		case []byte:
+			res.buffer = val
+		default:
+			err = fmt.Errorf("mssql: invalid type for varchar column: %s", val)
+			return
+		}
+		res.ti.Size = len(res.buffer)
+
+	case typeBit, typeBitN:
+		if reflect.TypeOf(val).Kind() != reflect.Bool {
+			err = fmt.Errorf("mssql: invalid type for bit column: %s", val)
+			return
+		}
+		res.ti.TypeId = typeBitN
+		res.ti.Size = 1
+		res.buffer = make([]byte, 1)
+		if val.(bool) {
+			res.buffer[0] = 1
+		}
+
+	case typeDateTime2N, typeDateTimeOffsetN:
+		switch val := val.(type) {
+		case time.Time:
+			days, ns := dateTime2(val)
+			ns /= int64(math.Pow10(int(col.ti.Scale)*-1) * 1000000000)
+
+			var data = make([]byte, 5)
+
+			data[0] = byte(ns)
+			data[1] = byte(ns >> 8)
+			data[2] = byte(ns >> 16)
+			data[3] = byte(ns >> 24)
+			data[4] = byte(ns >> 32)
+
+			if col.ti.Scale <= 2 {
+				res.ti.Size = 6
+			} else if col.ti.Scale <= 4 {
+				res.ti.Size = 7
+			} else {
+				res.ti.Size = 8
+			}
+			var buf []byte
+			buf = make([]byte, res.ti.Size)
+			copy(buf, data[0:res.ti.Size-3])
+
+			buf[res.ti.Size-3] = byte(days)
+			buf[res.ti.Size-2] = byte(days >> 8)
+			buf[res.ti.Size-1] = byte(days >> 16)
+
+			if col.ti.TypeId == typeDateTimeOffsetN {
+				_, offset := val.Zone()
+				var offsetMinute = uint16(offset / 60)
+				buf = append(buf, byte(offsetMinute))
+				buf = append(buf, byte(offsetMinute>>8))
+				res.ti.Size = res.ti.Size + 2
+			}
+
+			res.buffer = buf
+
+		default:
+			err = fmt.Errorf("mssql: invalid type for datetime2 column: %s", val)
+			return
+		}
+	case typeDateN:
+		switch val := val.(type) {
+		case time.Time:
+			days, _ := dateTime2(val)
+
+			res.ti.Size = 3
+			res.buffer = make([]byte, 3)
+			res.buffer[0] = byte(days)
+			res.buffer[1] = byte(days >> 8)
+			res.buffer[2] = byte(days >> 16)
+		default:
+			err = fmt.Errorf("mssql: invalid type for date column: %s", val)
+			return
+		}
+	case typeDateTime, typeDateTimeN, typeDateTim4:
+		switch val := val.(type) {
+		case time.Time:
+			if col.ti.Size == 4 {
+				res.ti.Size = 4
+				res.buffer = make([]byte, 4)
+
+				ref := time.Date(1900, 1, 1, 0, 0, 0, 0, time.UTC)
+				dur := val.Sub(ref)
+				days := dur / (24 * time.Hour)
+				if days < 0 {
+					err = fmt.Errorf("mssql: Date %s is out of range", val)
+					return
+				}
+				mins := val.Hour()*60 + val.Minute()
+
+				binary.LittleEndian.PutUint16(res.buffer[0:2], uint16(days))
+				binary.LittleEndian.PutUint16(res.buffer[2:4], uint16(mins))
+			} else if col.ti.Size == 8 {
+				res.ti.Size = 8
+				res.buffer = make([]byte, 8)
+
+				days := divFloor(val.Unix(), 24*60*60)
+				//25567 - number of days since Jan 1 1900 UTC to Jan 1 1970
+				days = days + 25567
+				tm := (val.Hour()*60*60+val.Minute()*60+val.Second())*300 + int(val.Nanosecond()/10000000*3)
+
+				binary.LittleEndian.PutUint32(res.buffer[0:4], uint32(days))
+				binary.LittleEndian.PutUint32(res.buffer[4:8], uint32(tm))
+			} else {
+				err = fmt.Errorf("mssql: invalid size of column")
+			}
+
+		default:
+			err = fmt.Errorf("mssql: invalid type for datetime column: %s", val)
+		}
+
+	// case typeMoney, typeMoney4, typeMoneyN:
+	case typeDecimal, typeDecimalN, typeNumeric, typeNumericN:
+		var value float64
+		switch v := val.(type) {
+		case int:
+			value = float64(v)
+		case int8:
+			value = float64(v)
+		case int16:
+			value = float64(v)
+		case int32:
+			value = float64(v)
+		case int64:
+			value = float64(v)
+		case float32:
+			value = float64(v)
+		case float64:
+			value = v
+		case string:
+			if value, err = strconv.ParseFloat(v, 64); err != nil {
+				return res, fmt.Errorf("bulk: unable to convert string to float: %v", err)
+			}
+		default:
+			return res, fmt.Errorf("unknown value for decimal: %#v", v)
+		}
+
+		perc := col.ti.Prec
+		scale := col.ti.Scale
+		var dec Decimal
+		dec, err = Float64ToDecimalScale(value, scale)
+		if err != nil {
+			return res, err
+		}
+		dec.prec = perc
+
+		var length byte
+		switch {
+		case perc <= 9:
+			length = 4
+		case perc <= 19:
+			length = 8
+		case perc <= 28:
+			length = 12
+		default:
+			length = 16
+		}
+
+		buf := make([]byte, length+1)
+		// first byte length written by typeInfo.writer
+		res.ti.Size = int(length) + 1
+		// second byte sign
+		if value < 0 {
+			buf[0] = 0
+		} else {
+			buf[0] = 1
+		}
+
+		ub := dec.UnscaledBytes()
+		l := len(ub)
+		if l > int(length) {
+			err = fmt.Errorf("decimal out of range: %s", dec)
+			return res, err
+		}
+		// reverse the bytes
+		for i, j := 1, l-1; j >= 0; i, j = i+1, j-1 {
+			buf[i] = ub[j]
+		}
+		res.buffer = buf
+	case typeBigVarBin:
+		switch val := val.(type) {
+		case []byte:
+			res.ti.Size = len(val)
+			res.buffer = val
+		default:
+			err = fmt.Errorf("mssql: invalid type for Binary column: %s", val)
+			return
+		}
+	case typeGuid:
+		switch val := val.(type) {
+		case []byte:
+			res.ti.Size = len(val)
+			res.buffer = val
+		default:
+			err = fmt.Errorf("mssql: invalid type for Guid column: %s", val)
+			return
+		}
+
+	default:
+		err = fmt.Errorf("mssql: type %x not implemented", col.ti.TypeId)
+	}
+	return
+
+}
+
+func (b *Bulk) dlogf(format string, v ...interface{}) {
+	if b.Debug {
+		b.cn.sess.log.Printf(format, v...)
+	}
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/bulkcopy_sql.go b/vendor/github.com/denisenkom/go-mssqldb/bulkcopy_sql.go
new file mode 100644
index 0000000000..0af51df890
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/bulkcopy_sql.go
@@ -0,0 +1,93 @@
+package mssql
+
+import (
+	"context"
+	"database/sql/driver"
+	"encoding/json"
+	"errors"
+)
+
+type copyin struct {
+	cn       *Conn
+	bulkcopy *Bulk
+	closed   bool
+}
+
+type serializableBulkConfig struct {
+	TableName   string
+	ColumnsName []string
+	Options     BulkOptions
+}
+
+func (d *Driver) OpenConnection(dsn string) (*Conn, error) {
+	return d.open(context.Background(), dsn)
+}
+
+func (c *Conn) prepareCopyIn(query string) (_ driver.Stmt, err error) {
+	config_json := query[11:]
+
+	bulkconfig := serializableBulkConfig{}
+	err = json.Unmarshal([]byte(config_json), &bulkconfig)
+	if err != nil {
+		return
+	}
+
+	bulkcopy := c.CreateBulk(bulkconfig.TableName, bulkconfig.ColumnsName)
+	bulkcopy.Options = bulkconfig.Options
+
+	ci := &copyin{
+		cn:       c,
+		bulkcopy: bulkcopy,
+	}
+
+	return ci, nil
+}
+
+func CopyIn(table string, options BulkOptions, columns ...string) string {
+	bulkconfig := &serializableBulkConfig{TableName: table, Options: options, ColumnsName: columns}
+
+	config_json, err := json.Marshal(bulkconfig)
+	if err != nil {
+		panic(err)
+	}
+
+	stmt := "INSERTBULK " + string(config_json)
+
+	return stmt
+}
+
+func (ci *copyin) NumInput() int {
+	return -1
+}
+
+func (ci *copyin) Query(v []driver.Value) (r driver.Rows, err error) {
+	return nil, errors.New("ErrNotSupported")
+}
+
+func (ci *copyin) Exec(v []driver.Value) (r driver.Result, err error) {
+	if ci.closed {
+		return nil, errors.New("errCopyInClosed")
+	}
+
+	if len(v) == 0 {
+		rowCount, err := ci.bulkcopy.Done()
+		ci.closed = true
+		return driver.RowsAffected(rowCount), err
+	}
+
+	t := make([]interface{}, len(v))
+	for i, val := range v {
+		t[i] = val
+	}
+
+	err = ci.bulkcopy.AddRow(t)
+	if err != nil {
+		return
+	}
+
+	return driver.RowsAffected(0), nil
+}
+
+func (ci *copyin) Close() (err error) {
+	return nil
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/collation.go b/vendor/github.com/denisenkom/go-mssqldb/collation.go
deleted file mode 100644
index ac9cf20b7b..0000000000
--- a/vendor/github.com/denisenkom/go-mssqldb/collation.go
+++ /dev/null
@@ -1,39 +0,0 @@
-package mssql
-
-import (
-	"encoding/binary"
-	"io"
-)
-
-// http://msdn.microsoft.com/en-us/library/dd340437.aspx
-
-type collation struct {
-	lcidAndFlags uint32
-	sortId       uint8
-}
-
-func (c collation) getLcid() uint32 {
-	return c.lcidAndFlags & 0x000fffff
-}
-
-func (c collation) getFlags() uint32 {
-	return (c.lcidAndFlags & 0x0ff00000) >> 20
-}
-
-func (c collation) getVersion() uint32 {
-	return (c.lcidAndFlags & 0xf0000000) >> 28
-}
-
-func readCollation(r *tdsBuffer) (res collation) {
-	res.lcidAndFlags = r.uint32()
-	res.sortId = r.byte()
-	return
-}
-
-func writeCollation(w io.Writer, col collation) (err error) {
-	if err = binary.Write(w, binary.LittleEndian, col.lcidAndFlags); err != nil {
-		return
-	}
-	err = binary.Write(w, binary.LittleEndian, col.sortId)
-	return
-}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/decimal.go b/vendor/github.com/denisenkom/go-mssqldb/decimal.go
index 76f3a6b5b4..372f64b4eb 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/decimal.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/decimal.go
@@ -32,7 +32,13 @@ func (d Decimal) ToFloat64() float64 {
 	return val
 }
 
+const autoScale = 100
+
 func Float64ToDecimal(f float64) (Decimal, error) {
+	return Float64ToDecimalScale(f, autoScale)
+}
+
+func Float64ToDecimalScale(f float64, scale uint8) (Decimal, error) {
 	var dec Decimal
 	if math.IsNaN(f) {
 		return dec, errors.New("NaN")
@@ -49,10 +55,10 @@ func Float64ToDecimal(f float64) (Decimal, error) {
 	}
 	dec.prec = 20
 	var integer float64
-	for dec.scale = 0; dec.scale <= 20; dec.scale++ {
+	for dec.scale = 0; dec.scale <= scale; dec.scale++ {
 		integer = f * scaletblflt64[dec.scale]
 		_, frac := math.Modf(integer)
-		if frac == 0 {
+		if frac == 0 && scale == autoScale {
 			break
 		}
 	}
@@ -73,7 +79,7 @@ func init() {
 	}
 }
 
-func (d Decimal) Bytes() []byte {
+func (d Decimal) BigInt() big.Int {
 	bytes := make([]byte, 16)
 	binary.BigEndian.PutUint32(bytes[0:4], d.integer[3])
 	binary.BigEndian.PutUint32(bytes[4:8], d.integer[2])
@@ -84,9 +90,19 @@ func (d Decimal) Bytes() []byte {
 	if !d.positive {
 		x.Neg(&x)
 	}
+	return x
+}
+
+func (d Decimal) Bytes() []byte {
+	x := d.BigInt()
 	return scaleBytes(x.String(), d.scale)
 }
 
+func (d Decimal) UnscaledBytes() []byte {
+	x := d.BigInt()
+	return x.Bytes()
+}
+
 func scaleBytes(s string, scale uint8) []byte {
 	z := make([]byte, 0, len(s)+1)
 	if s[0] == '-' || s[0] == '+' {
diff --git a/vendor/github.com/denisenkom/go-mssqldb/doc.go b/vendor/github.com/denisenkom/go-mssqldb/doc.go
new file mode 100644
index 0000000000..1bb80c442b
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/doc.go
@@ -0,0 +1,12 @@
+// package mssql implements the TDS protocol used to connect to MS SQL Server (sqlserver)
+// database servers.
+//
+// This package registers two drivers:
+//    sqlserver: uses native "@" parameter placeholder names and does no pre-processing.
+//    mssql: expects identifiers to be prefixed with ":" and pre-processes queries.
+//
+// If the ordinal position is used for query parameters, identifiers will be named
+// "@p1", "@p2", ... "@pN".
+//
+// Please refer to the README for the format of the DSN.
+package mssql
diff --git a/vendor/github.com/denisenkom/go-mssqldb/charset.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/charset.go
similarity index 94%
rename from vendor/github.com/denisenkom/go-mssqldb/charset.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/charset.go
index f1cc247a9d..8dc2279ea4 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/charset.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/charset.go
@@ -1,14 +1,14 @@
-package mssql
+package cp
 
 type charsetMap struct {
 	sb [256]rune    // single byte runes, -1 for a double byte character lead byte
 	db map[int]rune // double byte runes
 }
 
-func collation2charset(col collation) *charsetMap {
+func collation2charset(col Collation) *charsetMap {
 	// http://msdn.microsoft.com/en-us/library/ms144250.aspx
 	// http://msdn.microsoft.com/en-us/library/ms144250(v=sql.105).aspx
-	switch col.sortId {
+	switch col.SortId {
 	case 30, 31, 32, 33, 34:
 		return cp437
 	case 40, 41, 42, 44, 49, 55, 56, 57, 58, 59, 60, 61:
@@ -86,7 +86,7 @@ func collation2charset(col collation) *charsetMap {
 	return cp1252
 }
 
-func charset2utf8(col collation, s []byte) string {
+func CharsetToUTF8(col Collation, s []byte) string {
 	cm := collation2charset(col)
 	if cm == nil {
 		return string(s)
diff --git a/vendor/github.com/denisenkom/go-mssqldb/internal/cp/collation.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/collation.go
new file mode 100644
index 0000000000..ae7b03bf13
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/collation.go
@@ -0,0 +1,20 @@
+package cp
+
+// http://msdn.microsoft.com/en-us/library/dd340437.aspx
+
+type Collation struct {
+	LcidAndFlags uint32
+	SortId       uint8
+}
+
+func (c Collation) getLcid() uint32 {
+	return c.LcidAndFlags & 0x000fffff
+}
+
+func (c Collation) getFlags() uint32 {
+	return (c.LcidAndFlags & 0x0ff00000) >> 20
+}
+
+func (c Collation) getVersion() uint32 {
+	return (c.LcidAndFlags & 0xf0000000) >> 28
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1250.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1250.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1250.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1250.go
index 8207366be7..5c8094ec3c 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1250.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1250.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1250 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1251.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1251.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1251.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1251.go
index f5b81c3934..dc5896770c 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1251.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1251.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1251 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1252.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1252.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1252.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1252.go
index ed705d35a7..5ae8703542 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1252.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1252.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1252 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1253.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1253.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1253.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1253.go
index cb1e1a7623..52c8e07aa6 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1253.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1253.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1253 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1254.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1254.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1254.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1254.go
index a4b09bb44f..5d8864a521 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1254.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1254.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1254 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1255.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1255.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1255.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1255.go
index 97f9ee9e91..60619895d9 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1255.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1255.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1255 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1256.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1256.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1256.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1256.go
index e91241b448..ffd04b3e5b 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1256.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1256.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1256 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1257.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1257.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1257.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1257.go
index bd93e6f891..492da72ea4 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1257.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1257.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1257 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp1258.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1258.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp1258.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1258.go
index 4e1f8ac943..80be52c596 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp1258.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp1258.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp1258 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp437.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp437.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp437.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp437.go
index f47f8ecc77..76dedfb8ef 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp437.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp437.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp437 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp850.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp850.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp850.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp850.go
index e6b3d16904..927ab249ef 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp850.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp850.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp850 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp874.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp874.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp874.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp874.go
index 9d691a1a59..723bf6c392 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp874.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp874.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp874 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp932.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp932.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp932.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp932.go
index 980c55d815..5fc1377424 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp932.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp932.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp932 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp936.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp936.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp936.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp936.go
index fca5da76d4..d1fac12e26 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp936.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp936.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp936 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp949.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp949.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp949.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp949.go
index cddfcbc852..52c708dfa5 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp949.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp949.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp949 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/cp950.go b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp950.go
similarity index 99%
rename from vendor/github.com/denisenkom/go-mssqldb/cp950.go
rename to vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp950.go
index cbf25cb91a..1301cd0f05 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/cp950.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/internal/cp/cp950.go
@@ -1,4 +1,4 @@
-package mssql
+package cp
 
 var cp950 *charsetMap = &charsetMap{
 	sb: [256]rune{
diff --git a/vendor/github.com/denisenkom/go-mssqldb/log.go b/vendor/github.com/denisenkom/go-mssqldb/log.go
index f350aed099..9b8c551e88 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/log.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/log.go
@@ -4,19 +4,26 @@ import (
 	"log"
 )
 
-type Logger log.Logger
+type Logger interface {
+	Printf(format string, v ...interface{})
+	Println(v ...interface{})
+}
 
-func (logger *Logger) Printf(format string, v ...interface{}) {
-	if logger != nil {
-		(*log.Logger)(logger).Printf(format, v...)
+type optionalLogger struct {
+	logger Logger
+}
+
+func (o optionalLogger) Printf(format string, v ...interface{}) {
+	if o.logger != nil {
+		o.logger.Printf(format, v...)
 	} else {
 		log.Printf(format, v...)
 	}
 }
 
-func (logger *Logger) Println(v ...interface{}) {
-	if logger != nil {
-		(*log.Logger)(logger).Println(v...)
+func (o optionalLogger) Println(v ...interface{}) {
+	if o.logger != nil {
+		o.logger.Println(v...)
 	} else {
 		log.Println(v...)
 	}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/mssql.go b/vendor/github.com/denisenkom/go-mssqldb/mssql.go
index 9663651e7c..8f5ff2d0ce 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/mssql.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/mssql.go
@@ -1,122 +1,269 @@
 package mssql
 
 import (
+	"context"
 	"database/sql"
 	"database/sql/driver"
 	"encoding/binary"
 	"errors"
 	"fmt"
 	"io"
-	"log"
 	"math"
 	"net"
+	"reflect"
 	"strings"
 	"time"
 )
 
+var driverInstance = &Driver{processQueryText: true}
+var driverInstanceNoProcess = &Driver{processQueryText: false}
+
 func init() {
-	sql.Register("mssql", &MssqlDriver{})
+	sql.Register("mssql", driverInstance)
+	sql.Register("sqlserver", driverInstanceNoProcess)
+	createDialer = func(p *connectParams) dialer {
+		return tcpDialer{&net.Dialer{Timeout: p.dial_timeout, KeepAlive: p.keepAlive}}
+	}
 }
 
-type MssqlDriver struct {
-	log *log.Logger
+// Abstract the dialer for testing and for non-TCP based connections.
+type dialer interface {
+	Dial(ctx context.Context, addr string) (net.Conn, error)
 }
 
-func (d *MssqlDriver) SetLogger(logger *log.Logger) {
-	d.log = logger
+var createDialer func(p *connectParams) dialer
+
+type tcpDialer struct {
+	nd *net.Dialer
 }
 
-func CheckBadConn(err error) error {
-	if err == io.EOF {
+func (d tcpDialer) Dial(ctx context.Context, addr string) (net.Conn, error) {
+	return d.nd.DialContext(ctx, "tcp", addr)
+}
+
+type Driver struct {
+	log optionalLogger
+
+	processQueryText bool
+}
+
+// OpenConnector opens a new connector. Useful to dial with a context.
+func (d *Driver) OpenConnector(dsn string) (*Connector, error) {
+	params, err := parseConnectParams(dsn)
+	if err != nil {
+		return nil, err
+	}
+	return &Connector{
+		params: params,
+		driver: d,
+	}, nil
+}
+
+func (d *Driver) Open(dsn string) (driver.Conn, error) {
+	return d.open(context.Background(), dsn)
+}
+
+// Connector holds the parsed DSN and is ready to make a new connection
+// at any time.
+//
+// In the future, settings that cannot be passed through a string DSN
+// may be set directly on the connector.
+type Connector struct {
+	params connectParams
+	driver *Driver
+}
+
+// Connect to the server and return a TDS connection.
+func (c *Connector) Connect(ctx context.Context) (driver.Conn, error) {
+	return c.driver.connect(ctx, c.params)
+}
+
+// Driver underlying the Connector.
+func (c *Connector) Driver() driver.Driver {
+	return c.driver
+}
+
+func SetLogger(logger Logger) {
+	driverInstance.SetLogger(logger)
+	driverInstanceNoProcess.SetLogger(logger)
+}
+
+func (d *Driver) SetLogger(logger Logger) {
+	d.log = optionalLogger{logger}
+}
+
+type Conn struct {
+	sess           *tdsSession
+	transactionCtx context.Context
+
+	processQueryText bool
+	connectionGood   bool
+
+	outs map[string]interface{}
+}
+
+func (c *Conn) checkBadConn(err error) error {
+	// this is a hack to address Issue #275
+	// we set connectionGood flag to false if
+	// error indicates that connection is not usable
+	// but we return actual error instead of ErrBadConn
+	// this will cause connection to stay in a pool
+	// but next request to this connection will return ErrBadConn
+
+	// it might be possible to revise this hack after
+	// https://github.com/golang/go/issues/20807
+	// is implemented
+	switch err {
+	case nil:
+		return nil
+	case io.EOF:
 		return driver.ErrBadConn
+	case driver.ErrBadConn:
+		// It is an internal programming error if driver.ErrBadConn
+		// is ever passed to this function. driver.ErrBadConn should
+		// only ever be returned in response to a *mssql.Conn.connectionGood == false
+		// check in the external facing API.
+		panic("driver.ErrBadConn in checkBadConn. This should not happen.")
 	}
 
-	switch e := err.(type) {
+	switch err.(type) {
 	case net.Error:
-		if e.Timeout() {
-			return e
-		}
-		return driver.ErrBadConn
+		c.connectionGood = false
+		return err
+	case StreamError:
+		c.connectionGood = false
+		return err
 	default:
 		return err
 	}
 }
 
-type MssqlConn struct {
-	sess *tdsSession
+func (c *Conn) clearOuts() {
+	c.outs = nil
 }
 
-func (c *MssqlConn) Commit() error {
+func (c *Conn) simpleProcessResp(ctx context.Context) error {
+	tokchan := make(chan tokenStruct, 5)
+	go processResponse(ctx, c.sess, tokchan, c.outs)
+	c.clearOuts()
+	for tok := range tokchan {
+		switch token := tok.(type) {
+		case doneStruct:
+			if token.isError() {
+				return c.checkBadConn(token.getError())
+			}
+		case error:
+			return c.checkBadConn(token)
+		}
+	}
+	return nil
+}
+
+func (c *Conn) Commit() error {
+	if !c.connectionGood {
+		return driver.ErrBadConn
+	}
+	if err := c.sendCommitRequest(); err != nil {
+		return c.checkBadConn(err)
+	}
+	return c.simpleProcessResp(c.transactionCtx)
+}
+
+func (c *Conn) sendCommitRequest() error {
 	headers := []headerStruct{
 		{hdrtype: dataStmHdrTransDescr,
 			data: transDescrHdr{c.sess.tranid, 1}.pack()},
 	}
 	if err := sendCommitXact(c.sess.buf, headers, "", 0, 0, ""); err != nil {
-		return err
-	}
-
-	tokchan := make(chan tokenStruct, 5)
-	go processResponse(c.sess, tokchan)
-	for tok := range tokchan {
-		switch token := tok.(type) {
-		case error:
-			return token
+		if c.sess.logFlags&logErrors != 0 {
+			c.sess.log.Printf("Failed to send CommitXact with %v", err)
 		}
+		c.connectionGood = false
+		return fmt.Errorf("Faild to send CommitXact: %v", err)
 	}
 	return nil
 }
 
-func (c *MssqlConn) Rollback() error {
+func (c *Conn) Rollback() error {
+	if !c.connectionGood {
+		return driver.ErrBadConn
+	}
+	if err := c.sendRollbackRequest(); err != nil {
+		return c.checkBadConn(err)
+	}
+	return c.simpleProcessResp(c.transactionCtx)
+}
+
+func (c *Conn) sendRollbackRequest() error {
 	headers := []headerStruct{
 		{hdrtype: dataStmHdrTransDescr,
 			data: transDescrHdr{c.sess.tranid, 1}.pack()},
 	}
 	if err := sendRollbackXact(c.sess.buf, headers, "", 0, 0, ""); err != nil {
-		return err
-	}
-
-	tokchan := make(chan tokenStruct, 5)
-	go processResponse(c.sess, tokchan)
-	for tok := range tokchan {
-		switch token := tok.(type) {
-		case error:
-			return token
+		if c.sess.logFlags&logErrors != 0 {
+			c.sess.log.Printf("Failed to send RollbackXact with %v", err)
 		}
+		c.connectionGood = false
+		return fmt.Errorf("Failed to send RollbackXact: %v", err)
 	}
 	return nil
 }
 
-func (c *MssqlConn) Begin() (driver.Tx, error) {
+func (c *Conn) Begin() (driver.Tx, error) {
+	return c.begin(context.Background(), isolationUseCurrent)
+}
+
+func (c *Conn) begin(ctx context.Context, tdsIsolation isoLevel) (tx driver.Tx, err error) {
+	if !c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	err = c.sendBeginRequest(ctx, tdsIsolation)
+	if err != nil {
+		return nil, c.checkBadConn(err)
+	}
+	tx, err = c.processBeginResponse(ctx)
+	if err != nil {
+		return nil, c.checkBadConn(err)
+	}
+	return
+}
+
+func (c *Conn) sendBeginRequest(ctx context.Context, tdsIsolation isoLevel) error {
+	c.transactionCtx = ctx
 	headers := []headerStruct{
 		{hdrtype: dataStmHdrTransDescr,
 			data: transDescrHdr{0, 1}.pack()},
 	}
-	if err := sendBeginXact(c.sess.buf, headers, 0, ""); err != nil {
-		return nil, CheckBadConn(err)
-	}
-	tokchan := make(chan tokenStruct, 5)
-	go processResponse(c.sess, tokchan)
-	for tok := range tokchan {
-		switch token := tok.(type) {
-		case error:
-			if c.sess.tranid != 0 {
-				return nil, token
-			}
-			return nil, CheckBadConn(token)
+	if err := sendBeginXact(c.sess.buf, headers, tdsIsolation, ""); err != nil {
+		if c.sess.logFlags&logErrors != 0 {
+			c.sess.log.Printf("Failed to send BeginXact with %v", err)
 		}
+		c.connectionGood = false
+		return fmt.Errorf("Failed to send BiginXant: %v", err)
+	}
+	return nil
+}
+
+func (c *Conn) processBeginResponse(ctx context.Context) (driver.Tx, error) {
+	if err := c.simpleProcessResp(ctx); err != nil {
+		return nil, err
 	}
 	// successful BEGINXACT request will return sess.tranid
 	// for started transaction
 	return c, nil
 }
 
-func (d *MssqlDriver) Open(dsn string) (driver.Conn, error) {
+func (d *Driver) open(ctx context.Context, dsn string) (*Conn, error) {
 	params, err := parseConnectParams(dsn)
 	if err != nil {
 		return nil, err
 	}
+	return d.connect(ctx, params)
+}
 
-	sess, err := connect(params)
+// connect to the server, using the provided context for dialing only.
+func (d *Driver) connect(ctx context.Context, params connectParams) (*Conn, error) {
+	sess, err := connect(ctx, d.log, params)
 	if err != nil {
 		// main server failed, try fail-over partner
 		if params.failOverPartner == "" {
@@ -128,24 +275,29 @@ func (d *MssqlDriver) Open(dsn string) (driver.Conn, error) {
 			params.port = params.failOverPort
 		}
 
-		sess, err = connect(params)
+		sess, err = connect(ctx, d.log, params)
 		if err != nil {
 			// fail-over partner also failed, now fail
 			return nil, err
 		}
 	}
 
-	conn := &MssqlConn{sess}
-	conn.sess.log = (*Logger)(d.log)
+	conn := &Conn{
+		sess:             sess,
+		transactionCtx:   context.Background(),
+		processQueryText: d.processQueryText,
+		connectionGood:   true,
+	}
+	conn.sess.log = d.log
 	return conn, nil
 }
 
-func (c *MssqlConn) Close() error {
+func (c *Conn) Close() error {
 	return c.sess.buf.transport.Close()
 }
 
-type MssqlStmt struct {
-	c          *MssqlConn
+type Stmt struct {
+	c          *Conn
 	query      string
 	paramCount int
 	notifSub   *queryNotifSub
@@ -157,16 +309,30 @@ type queryNotifSub struct {
 	timeout uint32
 }
 
-func (c *MssqlConn) Prepare(query string) (driver.Stmt, error) {
-	q, paramCount := parseParams(query)
-	return &MssqlStmt{c, q, paramCount, nil}, nil
+func (c *Conn) Prepare(query string) (driver.Stmt, error) {
+	if !c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	if len(query) > 10 && strings.EqualFold(query[:10], "INSERTBULK") {
+		return c.prepareCopyIn(query)
+	}
+
+	return c.prepareContext(context.Background(), query)
 }
 
-func (s *MssqlStmt) Close() error {
+func (c *Conn) prepareContext(ctx context.Context, query string) (*Stmt, error) {
+	paramCount := -1
+	if c.processQueryText {
+		query, paramCount = parseParams(query)
+	}
+	return &Stmt{c, query, paramCount, nil}, nil
+}
+
+func (s *Stmt) Close() error {
 	return nil
 }
 
-func (s *MssqlStmt) SetQueryNotification(id, options string, timeout time.Duration) {
+func (s *Stmt) SetQueryNotification(id, options string, timeout time.Duration) {
 	to := uint32(timeout / time.Second)
 	if to < 1 {
 		to = 1
@@ -174,183 +340,273 @@ func (s *MssqlStmt) SetQueryNotification(id, options string, timeout time.Durati
 	s.notifSub = &queryNotifSub{id, options, to}
 }
 
-func (s *MssqlStmt) NumInput() int {
+func (s *Stmt) NumInput() int {
 	return s.paramCount
 }
 
-func (s *MssqlStmt) sendQuery(args []driver.Value) (err error) {
+func (s *Stmt) sendQuery(args []namedValue) (err error) {
 	headers := []headerStruct{
 		{hdrtype: dataStmHdrTransDescr,
 			data: transDescrHdr{s.c.sess.tranid, 1}.pack()},
 	}
 
 	if s.notifSub != nil {
-		headers = append(headers, headerStruct{hdrtype: dataStmHdrQueryNotif,
-			data: queryNotifHdr{s.notifSub.msgText, s.notifSub.options, s.notifSub.timeout}.pack()})
+		headers = append(headers,
+			headerStruct{
+				hdrtype: dataStmHdrQueryNotif,
+				data: queryNotifHdr{
+					s.notifSub.msgText,
+					s.notifSub.options,
+					s.notifSub.timeout,
+				}.pack(),
+			})
 	}
 
-	if len(args) != s.paramCount {
-		return errors.New(fmt.Sprintf("sql: expected %d parameters, got %d", s.paramCount, len(args)))
-	}
+	// no need to check number of parameters here, it is checked by database/sql
 	if s.c.sess.logFlags&logSQL != 0 {
 		s.c.sess.log.Println(s.query)
 	}
 	if s.c.sess.logFlags&logParams != 0 && len(args) > 0 {
 		for i := 0; i < len(args); i++ {
-			s.c.sess.log.Printf("\t@p%d\t%v\n", i+1, args[i])
+			if len(args[i].Name) > 0 {
+				s.c.sess.log.Printf("\t@%s\t%v\n", args[i].Name, args[i].Value)
+			} else {
+				s.c.sess.log.Printf("\t@p%d\t%v\n", i+1, args[i].Value)
+			}
 		}
 
 	}
 	if len(args) == 0 {
 		if err = sendSqlBatch72(s.c.sess.buf, s.query, headers); err != nil {
-			if s.c.sess.tranid != 0 {
-				return err
+			if s.c.sess.logFlags&logErrors != 0 {
+				s.c.sess.log.Printf("Failed to send SqlBatch with %v", err)
 			}
-			return CheckBadConn(err)
+			s.c.connectionGood = false
+			return fmt.Errorf("failed to send SQL Batch: %v", err)
 		}
 	} else {
-		params := make([]Param, len(args)+2)
-		decls := make([]string, len(args))
-		params[0], err = s.makeParam(s.query)
-		if err != nil {
-			return
-		}
-		for i, val := range args {
-			params[i+2], err = s.makeParam(val)
+		proc := Sp_ExecuteSql
+		var params []Param
+		if isProc(s.query) {
+			proc.name = s.query
+			params, _, err = s.makeRPCParams(args, 0)
+		} else {
+			var decls []string
+			params, decls, err = s.makeRPCParams(args, 2)
 			if err != nil {
 				return
 			}
-			name := fmt.Sprintf("@p%d", i+1)
-			params[i+2].Name = name
-			decls[i] = fmt.Sprintf("%s %s", name, makeDecl(params[i+2].ti))
+			params[0] = makeStrParam(s.query)
+			params[1] = makeStrParam(strings.Join(decls, ","))
 		}
-		params[1], err = s.makeParam(strings.Join(decls, ","))
-		if err != nil {
-			return
-		}
-		if err = sendRpc(s.c.sess.buf, headers, Sp_ExecuteSql, 0, params); err != nil {
-			if s.c.sess.tranid != 0 {
-				return err
+		if err = sendRpc(s.c.sess.buf, headers, proc, 0, params); err != nil {
+			if s.c.sess.logFlags&logErrors != 0 {
+				s.c.sess.log.Printf("Failed to send Rpc with %v", err)
 			}
-			return CheckBadConn(err)
+			s.c.connectionGood = false
+			return fmt.Errorf("Failed to send RPC: %v", err)
 		}
 	}
 	return
 }
 
-func (s *MssqlStmt) Query(args []driver.Value) (res driver.Rows, err error) {
-	if err = s.sendQuery(args); err != nil {
-		return
+// isProc takes the query text in s and determines if it is a stored proc name
+// or SQL text.
+func isProc(s string) bool {
+	if len(s) == 0 {
+		return false
 	}
+	if s[0] == '[' && s[len(s)-1] == ']' && strings.ContainsAny(s, "\n\r") == false {
+		return true
+	}
+	return !strings.ContainsAny(s, " \t\n\r;")
+}
+
+func (s *Stmt) makeRPCParams(args []namedValue, offset int) ([]Param, []string, error) {
+	var err error
+	params := make([]Param, len(args)+offset)
+	decls := make([]string, len(args))
+	for i, val := range args {
+		params[i+offset], err = s.makeParam(val.Value)
+		if err != nil {
+			return nil, nil, err
+		}
+		var name string
+		if len(val.Name) > 0 {
+			name = "@" + val.Name
+		} else {
+			name = fmt.Sprintf("@p%d", val.Ordinal)
+		}
+		params[i+offset].Name = name
+		decls[i] = fmt.Sprintf("%s %s", name, makeDecl(params[i+offset].ti))
+	}
+	return params, decls, nil
+}
+
+type namedValue struct {
+	Name    string
+	Ordinal int
+	Value   driver.Value
+}
+
+func convertOldArgs(args []driver.Value) []namedValue {
+	list := make([]namedValue, len(args))
+	for i, v := range args {
+		list[i] = namedValue{
+			Ordinal: i + 1,
+			Value:   v,
+		}
+	}
+	return list
+}
+
+func (s *Stmt) Query(args []driver.Value) (driver.Rows, error) {
+	return s.queryContext(context.Background(), convertOldArgs(args))
+}
+
+func (s *Stmt) queryContext(ctx context.Context, args []namedValue) (rows driver.Rows, err error) {
+	if !s.c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	if err = s.sendQuery(args); err != nil {
+		return nil, s.c.checkBadConn(err)
+	}
+	return s.processQueryResponse(ctx)
+}
+
+func (s *Stmt) processQueryResponse(ctx context.Context) (res driver.Rows, err error) {
 	tokchan := make(chan tokenStruct, 5)
-	go processResponse(s.c.sess, tokchan)
+	ctx, cancel := context.WithCancel(ctx)
+	go processResponse(ctx, s.c.sess, tokchan, s.c.outs)
+	s.c.clearOuts()
 	// process metadata
-	var cols []string
+	var cols []columnStruct
 loop:
 	for tok := range tokchan {
 		switch token := tok.(type) {
-		// by ignoring DONE token we effectively
-		// skip empty result-sets
-		// this improves results in queryes like that:
+		// By ignoring DONE token we effectively
+		// skip empty result-sets.
+		// This improves results in queries like that:
 		// set nocount on; select 1
 		// see TestIgnoreEmptyResults test
 		//case doneStruct:
 		//break loop
 		case []columnStruct:
-			cols = make([]string, len(token))
-			for i, col := range token {
-				cols[i] = col.ColName
-			}
+			cols = token
 			break loop
-		case error:
-			if s.c.sess.tranid != 0 {
-				return nil, token
+		case doneStruct:
+			if token.isError() {
+				return nil, s.c.checkBadConn(token.getError())
 			}
-			return nil, CheckBadConn(token)
+		case error:
+			return nil, s.c.checkBadConn(token)
 		}
 	}
-	return &MssqlRows{sess: s.c.sess, tokchan: tokchan, cols: cols}, nil
+	res = &Rows{stmt: s, tokchan: tokchan, cols: cols, cancel: cancel}
+	return
 }
 
-func (s *MssqlStmt) Exec(args []driver.Value) (res driver.Result, err error) {
-	if err = s.sendQuery(args); err != nil {
-		return
+func (s *Stmt) Exec(args []driver.Value) (driver.Result, error) {
+	return s.exec(context.Background(), convertOldArgs(args))
+}
+
+func (s *Stmt) exec(ctx context.Context, args []namedValue) (res driver.Result, err error) {
+	if !s.c.connectionGood {
+		return nil, driver.ErrBadConn
 	}
+	if err = s.sendQuery(args); err != nil {
+		return nil, s.c.checkBadConn(err)
+	}
+	if res, err = s.processExec(ctx); err != nil {
+		return nil, s.c.checkBadConn(err)
+	}
+	return
+}
+
+func (s *Stmt) processExec(ctx context.Context) (res driver.Result, err error) {
 	tokchan := make(chan tokenStruct, 5)
-	go processResponse(s.c.sess, tokchan)
+	go processResponse(ctx, s.c.sess, tokchan, s.c.outs)
+	s.c.clearOuts()
 	var rowCount int64
 	for token := range tokchan {
 		switch token := token.(type) {
 		case doneInProcStruct:
 			if token.Status&doneCount != 0 {
-				rowCount = int64(token.RowCount)
+				rowCount += int64(token.RowCount)
 			}
 		case doneStruct:
 			if token.Status&doneCount != 0 {
-				rowCount = int64(token.RowCount)
+				rowCount += int64(token.RowCount)
+			}
+			if token.isError() {
+				return nil, token.getError()
 			}
 		case error:
-			if s.c.sess.logFlags&logErrors != 0 {
-				s.c.sess.log.Println("got error:", token)
-			}
-			if s.c.sess.tranid != 0 {
-				return nil, token
-			}
-			return nil, CheckBadConn(token)
+			return nil, token
 		}
 	}
-	return &MssqlResult{s.c, rowCount}, nil
+	return &Result{s.c, rowCount}, nil
 }
 
-type MssqlRows struct {
-	sess    *tdsSession
-	cols    []string
+type Rows struct {
+	stmt    *Stmt
+	cols    []columnStruct
 	tokchan chan tokenStruct
 
-	nextCols []string
+	nextCols []columnStruct
+
+	cancel func()
 }
 
-func (rc *MssqlRows) Close() error {
+func (rc *Rows) Close() error {
+	rc.cancel()
 	for _ = range rc.tokchan {
 	}
 	rc.tokchan = nil
 	return nil
 }
 
-func (rc *MssqlRows) Columns() (res []string) {
-	return rc.cols
+func (rc *Rows) Columns() (res []string) {
+	res = make([]string, len(rc.cols))
+	for i, col := range rc.cols {
+		res[i] = col.ColName
+	}
+	return
 }
 
-func (rc *MssqlRows) Next(dest []driver.Value) (err error) {
+func (rc *Rows) Next(dest []driver.Value) error {
+	if !rc.stmt.c.connectionGood {
+		return driver.ErrBadConn
+	}
 	if rc.nextCols != nil {
 		return io.EOF
 	}
 	for tok := range rc.tokchan {
 		switch tokdata := tok.(type) {
 		case []columnStruct:
-			cols := make([]string, len(tokdata))
-			for i, col := range tokdata {
-				cols[i] = col.ColName
-			}
-			rc.nextCols = cols
+			rc.nextCols = tokdata
 			return io.EOF
 		case []interface{}:
 			for i := range dest {
 				dest[i] = tokdata[i]
 			}
 			return nil
+		case doneStruct:
+			if tokdata.isError() {
+				return rc.stmt.c.checkBadConn(tokdata.getError())
+			}
 		case error:
-			return tokdata
+			return rc.stmt.c.checkBadConn(tokdata)
 		}
 	}
 	return io.EOF
 }
 
-func (rc *MssqlRows) HasNextResultSet() bool {
+func (rc *Rows) HasNextResultSet() bool {
 	return rc.nextCols != nil
 }
 
-func (rc *MssqlRows) NextResultSet() error {
+func (rc *Rows) NextResultSet() error {
 	rc.cols = rc.nextCols
 	rc.nextCols = nil
 	if rc.cols == nil {
@@ -359,11 +615,69 @@ func (rc *MssqlRows) NextResultSet() error {
 	return nil
 }
 
-func (s *MssqlStmt) makeParam(val driver.Value) (res Param, err error) {
+// It should return
+// the value type that can be used to scan types into. For example, the database
+// column type "bigint" this should return "reflect.TypeOf(int64(0))".
+func (r *Rows) ColumnTypeScanType(index int) reflect.Type {
+	return makeGoLangScanType(r.cols[index].ti)
+}
+
+// RowsColumnTypeDatabaseTypeName may be implemented by Rows. It should return the
+// database system type name without the length. Type names should be uppercase.
+// Examples of returned types: "VARCHAR", "NVARCHAR", "VARCHAR2", "CHAR", "TEXT",
+// "DECIMAL", "SMALLINT", "INT", "BIGINT", "BOOL", "[]BIGINT", "JSONB", "XML",
+// "TIMESTAMP".
+func (r *Rows) ColumnTypeDatabaseTypeName(index int) string {
+	return makeGoLangTypeName(r.cols[index].ti)
+}
+
+// RowsColumnTypeLength may be implemented by Rows. It should return the length
+// of the column type if the column is a variable length type. If the column is
+// not a variable length type ok should return false.
+// If length is not limited other than system limits, it should return math.MaxInt64.
+// The following are examples of returned values for various types:
+//   TEXT          (math.MaxInt64, true)
+//   varchar(10)   (10, true)
+//   nvarchar(10)  (10, true)
+//   decimal       (0, false)
+//   int           (0, false)
+//   bytea(30)     (30, true)
+func (r *Rows) ColumnTypeLength(index int) (int64, bool) {
+	return makeGoLangTypeLength(r.cols[index].ti)
+}
+
+// It should return
+// the precision and scale for decimal types. If not applicable, ok should be false.
+// The following are examples of returned values for various types:
+//   decimal(38, 4)    (38, 4, true)
+//   int               (0, 0, false)
+//   decimal           (math.MaxInt64, math.MaxInt64, true)
+func (r *Rows) ColumnTypePrecisionScale(index int) (int64, int64, bool) {
+	return makeGoLangTypePrecisionScale(r.cols[index].ti)
+}
+
+// The nullable value should
+// be true if it is known the column may be null, or false if the column is known
+// to be not nullable.
+// If the column nullability is unknown, ok should be false.
+func (r *Rows) ColumnTypeNullable(index int) (nullable, ok bool) {
+	nullable = r.cols[index].Flags&colFlagNullable != 0
+	ok = true
+	return
+}
+
+func makeStrParam(val string) (res Param) {
+	res.ti.TypeId = typeNVarChar
+	res.buffer = str2ucs2(val)
+	res.ti.Size = len(res.buffer)
+	return
+}
+
+func (s *Stmt) makeParam(val driver.Value) (res Param, err error) {
 	if val == nil {
-		res.ti.TypeId = typeNVarChar
+		res.ti.TypeId = typeNull
 		res.buffer = nil
-		res.ti.Size = 2
+		res.ti.Size = 0
 		return
 	}
 	switch val := val.(type) {
@@ -382,9 +696,7 @@ func (s *MssqlStmt) makeParam(val driver.Value) (res Param, err error) {
 		res.ti.Size = len(val)
 		res.buffer = val
 	case string:
-		res.ti.TypeId = typeNVarChar
-		res.buffer = str2ucs2(val)
-		res.ti.Size = len(res.buffer)
+		res = makeStrParam(val)
 	case bool:
 		res.ti.TypeId = typeBitN
 		res.ti.Size = 1
@@ -425,22 +737,21 @@ func (s *MssqlStmt) makeParam(val driver.Value) (res Param, err error) {
 			binary.LittleEndian.PutUint32(res.buffer[4:8], uint32(tm))
 		}
 	default:
-		err = fmt.Errorf("mssql: unknown type for %T", val)
-		return
+		return s.makeParamExtra(val)
 	}
 	return
 }
 
-type MssqlResult struct {
-	c            *MssqlConn
+type Result struct {
+	c            *Conn
 	rowsAffected int64
 }
 
-func (r *MssqlResult) RowsAffected() (int64, error) {
+func (r *Result) RowsAffected() (int64, error) {
 	return r.rowsAffected, nil
 }
 
-func (r *MssqlResult) LastInsertId() (int64, error) {
+func (r *Result) LastInsertId() (int64, error) {
 	s, err := r.c.Prepare("select cast(@@identity as bigint)")
 	if err != nil {
 		return 0, err
diff --git a/vendor/github.com/denisenkom/go-mssqldb/mssql_go1.3.go b/vendor/github.com/denisenkom/go-mssqldb/mssql_go1.3.go
deleted file mode 100644
index b8cffe9c01..0000000000
--- a/vendor/github.com/denisenkom/go-mssqldb/mssql_go1.3.go
+++ /dev/null
@@ -1,11 +0,0 @@
-// +build go1.3
-
-package mssql
-
-import (
-	"net"
-)
-
-func createDialer(p connectParams) *net.Dialer {
-	return &net.Dialer{Timeout: p.dial_timeout, KeepAlive: p.keepAlive}
-}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/mssql_go1.3pre.go b/vendor/github.com/denisenkom/go-mssqldb/mssql_go1.3pre.go
deleted file mode 100644
index 3c7e72716d..0000000000
--- a/vendor/github.com/denisenkom/go-mssqldb/mssql_go1.3pre.go
+++ /dev/null
@@ -1,11 +0,0 @@
-// +build !go1.3
-
-package mssql
-
-import (
-	"net"
-)
-
-func createDialer(p *connectParams) *net.Dialer {
-	return &net.Dialer{Timeout: p.dial_timeout}
-}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/mssql_go18.go b/vendor/github.com/denisenkom/go-mssqldb/mssql_go18.go
new file mode 100644
index 0000000000..74179c4720
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/mssql_go18.go
@@ -0,0 +1,91 @@
+// +build go1.8
+
+package mssql
+
+import (
+	"context"
+	"database/sql"
+	"database/sql/driver"
+	"errors"
+	"strings"
+)
+
+var _ driver.Pinger = &Conn{}
+
+// Ping is used to check if the remote server is available and satisfies the Pinger interface.
+func (c *Conn) Ping(ctx context.Context) error {
+	if !c.connectionGood {
+		return driver.ErrBadConn
+	}
+	stmt := &Stmt{c, `select 1;`, 0, nil}
+	_, err := stmt.ExecContext(ctx, nil)
+	return err
+}
+
+var _ driver.ConnBeginTx = &Conn{}
+
+// BeginTx satisfies ConnBeginTx.
+func (c *Conn) BeginTx(ctx context.Context, opts driver.TxOptions) (driver.Tx, error) {
+	if !c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	if opts.ReadOnly {
+		return nil, errors.New("Read-only transactions are not supported")
+	}
+
+	var tdsIsolation isoLevel
+	switch sql.IsolationLevel(opts.Isolation) {
+	case sql.LevelDefault:
+		tdsIsolation = isolationUseCurrent
+	case sql.LevelReadUncommitted:
+		tdsIsolation = isolationReadUncommited
+	case sql.LevelReadCommitted:
+		tdsIsolation = isolationReadCommited
+	case sql.LevelWriteCommitted:
+		return nil, errors.New("LevelWriteCommitted isolation level is not supported")
+	case sql.LevelRepeatableRead:
+		tdsIsolation = isolationRepeatableRead
+	case sql.LevelSnapshot:
+		tdsIsolation = isolationSnapshot
+	case sql.LevelSerializable:
+		tdsIsolation = isolationSerializable
+	case sql.LevelLinearizable:
+		return nil, errors.New("LevelLinearizable isolation level is not supported")
+	default:
+		return nil, errors.New("Isolation level is not supported or unknown")
+	}
+	return c.begin(ctx, tdsIsolation)
+}
+
+func (c *Conn) PrepareContext(ctx context.Context, query string) (driver.Stmt, error) {
+	if !c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	if len(query) > 10 && strings.EqualFold(query[:10], "INSERTBULK") {
+		return c.prepareCopyIn(query)
+	}
+
+	return c.prepareContext(ctx, query)
+}
+
+func (s *Stmt) QueryContext(ctx context.Context, args []driver.NamedValue) (driver.Rows, error) {
+	if !s.c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	list := make([]namedValue, len(args))
+	for i, nv := range args {
+		list[i] = namedValue(nv)
+	}
+	return s.queryContext(ctx, list)
+}
+
+func (s *Stmt) ExecContext(ctx context.Context, args []driver.NamedValue) (driver.Result, error) {
+	if !s.c.connectionGood {
+		return nil, driver.ErrBadConn
+	}
+	list := make([]namedValue, len(args))
+	for i, nv := range args {
+		list[i] = namedValue(nv)
+	}
+	return s.exec(ctx, list)
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/mssql_go19.go b/vendor/github.com/denisenkom/go-mssqldb/mssql_go19.go
new file mode 100644
index 0000000000..250151abc3
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/mssql_go19.go
@@ -0,0 +1,64 @@
+// +build go1.9
+
+package mssql
+
+import (
+	"database/sql"
+	"database/sql/driver"
+	"fmt"
+	// "github.com/cockroachdb/apd"
+)
+
+// Type alias provided for compibility.
+//
+// Deprecated: users should transition to the new names when possible.
+type MssqlDriver = Driver
+type MssqlBulk = Bulk
+type MssqlBulkOptions = BulkOptions
+type MssqlConn = Conn
+type MssqlResult = Result
+type MssqlRows = Rows
+type MssqlStmt = Stmt
+
+var _ driver.NamedValueChecker = &Conn{}
+
+func (c *Conn) CheckNamedValue(nv *driver.NamedValue) error {
+	switch v := nv.Value.(type) {
+	case sql.Out:
+		if c.outs == nil {
+			c.outs = make(map[string]interface{})
+		}
+		c.outs[nv.Name] = v.Dest
+
+		// Unwrap the Out value and check the inner value.
+		lnv := *nv
+		lnv.Value = v.Dest
+		err := c.CheckNamedValue(&lnv)
+		if err != nil {
+			if err != driver.ErrSkip {
+				return err
+			}
+			lnv.Value, err = driver.DefaultParameterConverter.ConvertValue(lnv.Value)
+			if err != nil {
+				return err
+			}
+		}
+		nv.Value = sql.Out{Dest: lnv.Value}
+		return nil
+	// case *apd.Decimal:
+	// 	return nil
+	default:
+		return driver.ErrSkip
+	}
+}
+
+func (s *Stmt) makeParamExtra(val driver.Value) (res Param, err error) {
+	switch val := val.(type) {
+	case sql.Out:
+		res, err = s.makeParam(val.Dest)
+		res.Flags = fByRevValue
+	default:
+		err = fmt.Errorf("mssql: unknown type for %T", val)
+	}
+	return
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/mssql_go19pre.go b/vendor/github.com/denisenkom/go-mssqldb/mssql_go19pre.go
new file mode 100644
index 0000000000..3bad7fb008
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/mssql_go19pre.go
@@ -0,0 +1,12 @@
+// +build !go1.9
+
+package mssql
+
+import (
+	"database/sql/driver"
+	"fmt"
+)
+
+func (s *Stmt) makeParamExtra(val driver.Value) (Param, error) {
+	return Param{}, fmt.Errorf("mssql: unknown type for %T", val)
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/net.go b/vendor/github.com/denisenkom/go-mssqldb/net.go
index 72a87340db..8c3c8ef8b1 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/net.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/net.go
@@ -33,7 +33,7 @@ func (c *timeoutConn) Read(b []byte) (n int, err error) {
 			c.continueRead = false
 		}
 		if !c.continueRead {
-			var packet uint8
+			var packet packetType
 			packet, err = c.buf.BeginRead()
 			if err != nil {
 				err = fmt.Errorf("Cannot read handshake packet: %s", err.Error())
diff --git a/vendor/github.com/denisenkom/go-mssqldb/ntlm.go b/vendor/github.com/denisenkom/go-mssqldb/ntlm.go
index f853435c6e..5bed668430 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/ntlm.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/ntlm.go
@@ -59,7 +59,7 @@ type NTLMAuth struct {
 	Workstation string
 }
 
-func getAuth(user, password, service, workstation string) (Auth, bool) {
+func getAuth(user, password, service, workstation string) (auth, bool) {
 	if !strings.ContainsRune(user, '\\') {
 		return nil, false
 	}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/parser.go b/vendor/github.com/denisenkom/go-mssqldb/parser.go
index 9e37c16a65..8021ca603c 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/parser.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/parser.go
@@ -11,6 +11,9 @@ type parser struct {
 	w          bytes.Buffer
 	paramCount int
 	paramMax   int
+
+	// using map as a set
+	namedParams map[string]bool
 }
 
 func (p *parser) next() (rune, bool) {
@@ -39,13 +42,14 @@ type stateFunc func(*parser) stateFunc
 
 func parseParams(query string) (string, int) {
 	p := &parser{
-		r: bytes.NewReader([]byte(query)),
+		r:           bytes.NewReader([]byte(query)),
+		namedParams: map[string]bool{},
 	}
 	state := parseNormal
 	for state != nil {
 		state = state(p)
 	}
-	return p.w.String(), p.paramMax
+	return p.w.String(), p.paramMax + len(p.namedParams)
 }
 
 func parseNormal(p *parser) stateFunc {
@@ -55,7 +59,7 @@ func parseNormal(p *parser) stateFunc {
 			return nil
 		}
 		if ch == '?' {
-			return parseParameter
+			return parseOrdinalParameter
 		} else if ch == '$' || ch == ':' {
 			ch2, ok := p.next()
 			if !ok {
@@ -64,7 +68,9 @@ func parseNormal(p *parser) stateFunc {
 			}
 			p.unread()
 			if ch2 >= '0' && ch2 <= '9' {
-				return parseParameter
+				return parseOrdinalParameter
+			} else if 'a' <= ch2 && ch2 <= 'z' || 'A' <= ch2 && ch2 <= 'Z' {
+				return parseNamedParameter
 			}
 		}
 		p.write(ch)
@@ -83,7 +89,7 @@ func parseNormal(p *parser) stateFunc {
 	}
 }
 
-func parseParameter(p *parser) stateFunc {
+func parseOrdinalParameter(p *parser) stateFunc {
 	var paramN int
 	var ok bool
 	for {
@@ -113,6 +119,30 @@ func parseParameter(p *parser) stateFunc {
 	return parseNormal
 }
 
+func parseNamedParameter(p *parser) stateFunc {
+	var paramName string
+	var ok bool
+	for {
+		var ch rune
+		ch, ok = p.next()
+		if ok && (ch >= '0' && ch <= '9' || 'a' <= ch && ch <= 'z' || 'A' <= ch && ch <= 'Z') {
+			paramName = paramName + string(ch)
+		} else {
+			break
+		}
+	}
+	if ok {
+		p.unread()
+	}
+	p.namedParams[paramName] = true
+	p.w.WriteString("@")
+	p.w.WriteString(paramName)
+	if !ok {
+		return nil
+	}
+	return parseNormal
+}
+
 func parseQuote(p *parser) stateFunc {
 	for {
 		ch, ok := p.next()
diff --git a/vendor/github.com/denisenkom/go-mssqldb/sspi_windows.go b/vendor/github.com/denisenkom/go-mssqldb/sspi_windows.go
index a6e95051c9..9b5bc6893f 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/sspi_windows.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/sspi_windows.go
@@ -113,7 +113,7 @@ type SSPIAuth struct {
 	ctxt     SecHandle
 }
 
-func getAuth(user, password, service, workstation string) (Auth, bool) {
+func getAuth(user, password, service, workstation string) (auth, bool) {
 	if user == "" {
 		return &SSPIAuth{Service: service}, true
 	}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/tds.go b/vendor/github.com/denisenkom/go-mssqldb/tds.go
index fd42dba34a..54ac6dbad5 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/tds.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/tds.go
@@ -1,6 +1,7 @@
 package mssql
 
 import (
+	"context"
 	"crypto/tls"
 	"crypto/x509"
 	"encoding/binary"
@@ -9,11 +10,13 @@ import (
 	"io"
 	"io/ioutil"
 	"net"
+	"net/url"
 	"os"
 	"sort"
 	"strconv"
 	"strings"
 	"time"
+	"unicode"
 	"unicode/utf16"
 	"unicode/utf8"
 )
@@ -47,8 +50,11 @@ func parseInstances(msg []byte) map[string]map[string]string {
 	return results
 }
 
-func getInstances(address string) (map[string]map[string]string, error) {
-	conn, err := net.DialTimeout("udp", address+":1434", 5*time.Second)
+func getInstances(ctx context.Context, address string) (map[string]map[string]string, error) {
+	dialer := &net.Dialer{
+		Timeout: 5 * time.Second,
+	}
+	conn, err := dialer.DialContext(ctx, "udp", address+":1434")
 	if err != nil {
 		return nil, err
 	}
@@ -79,11 +85,16 @@ const (
 )
 
 // packet types
+// https://msdn.microsoft.com/en-us/library/dd304214.aspx
 const (
-	packSQLBatch    = 1
-	packRPCRequest  = 3
-	packReply       = 4
-	packCancel      = 6
+	packSQLBatch   packetType = 1
+	packRPCRequest            = 3
+	packReply                 = 4
+
+	// 2.2.1.7 Attention: https://msdn.microsoft.com/en-us/library/dd341449.aspx
+	// 4.19.2 Out-of-Band Attention Signal: https://msdn.microsoft.com/en-us/library/dd305167.aspx
+	packAttention = 6
+
 	packBulkLoadBCP = 7
 	packTransMgrReq = 14
 	packNormal      = 15
@@ -119,7 +130,7 @@ type tdsSession struct {
 	columns      []columnStruct
 	tranid       uint64
 	logFlags     uint64
-	log          *Logger
+	log          optionalLogger
 	routedServer string
 	routedPort   uint16
 }
@@ -131,6 +142,7 @@ const (
 	logSQL         = 8
 	logParams      = 16
 	logTransaction = 32
+	logDebug       = 64
 )
 
 type columnStruct struct {
@@ -490,6 +502,11 @@ func readBVarChar(r io.Reader) (res string, err error) {
 	if err != nil {
 		return "", err
 	}
+
+	// A zero length could be returned, return an empty string
+	if numchars == 0 {
+		return "", nil
+	}
 	return readUcs2(r, int(numchars))
 }
 
@@ -588,7 +605,7 @@ func (hdr transDescrHdr) pack() (res []byte) {
 }
 
 func writeAllHeaders(w io.Writer, headers []headerStruct) (err error) {
-	// calculatint total length
+	// Calculating total length.
 	var totallen uint32 = 4
 	for _, hdr := range headers {
 		totallen += 4 + 2 + uint32(len(hdr.data))
@@ -616,9 +633,7 @@ func writeAllHeaders(w io.Writer, headers []headerStruct) (err error) {
 	return nil
 }
 
-func sendSqlBatch72(buf *tdsBuffer,
-	sqltext string,
-	headers []headerStruct) (err error) {
+func sendSqlBatch72(buf *tdsBuffer, sqltext string, headers []headerStruct) (err error) {
 	buf.BeginPacket(packSQLBatch)
 
 	if err = writeAllHeaders(buf, headers); err != nil {
@@ -632,6 +647,13 @@ func sendSqlBatch72(buf *tdsBuffer,
 	return buf.FinishPacket()
 }
 
+// 2.2.1.7 Attention: https://msdn.microsoft.com/en-us/library/dd341449.aspx
+// 4.19.2 Out-of-Band Attention Signal: https://msdn.microsoft.com/en-us/library/dd305167.aspx
+func sendAttention(buf *tdsBuffer) error {
+	buf.BeginPacket(packAttention)
+	return buf.FinishPacket()
+}
+
 type connectParams struct {
 	logFlags               uint64
 	port                   uint64
@@ -654,6 +676,7 @@ type connectParams struct {
 	typeFlags              uint8
 	failOverPartner        string
 	failOverPort           uint64
+	packetSize             uint16
 }
 
 func splitConnectionString(dsn string) (res map[string]string) {
@@ -677,9 +700,241 @@ func splitConnectionString(dsn string) (res map[string]string) {
 	return res
 }
 
+// Splits a URL in the ODBC format
+func splitConnectionStringOdbc(dsn string) (map[string]string, error) {
+	res := map[string]string{}
+
+	type parserState int
+	const (
+		// Before the start of a key
+		parserStateBeforeKey parserState = iota
+
+		// Inside a key
+		parserStateKey
+
+		// Beginning of a value. May be bare or braced
+		parserStateBeginValue
+
+		// Inside a bare value
+		parserStateBareValue
+
+		// Inside a braced value
+		parserStateBracedValue
+
+		// A closing brace inside a braced value.
+		// May be the end of the value or an escaped closing brace, depending on the next character
+		parserStateBracedValueClosingBrace
+
+		// After a value. Next character should be a semicolon or whitespace.
+		parserStateEndValue
+	)
+
+	var state = parserStateBeforeKey
+
+	var key string
+	var value string
+
+	for i, c := range dsn {
+		switch state {
+		case parserStateBeforeKey:
+			switch {
+			case c == '=':
+				return res, fmt.Errorf("Unexpected character = at index %d. Expected start of key or semi-colon or whitespace.", i)
+			case !unicode.IsSpace(c) && c != ';':
+				state = parserStateKey
+				key += string(c)
+			}
+
+		case parserStateKey:
+			switch c {
+			case '=':
+				key = normalizeOdbcKey(key)
+				if len(key) == 0 {
+					return res, fmt.Errorf("Unexpected end of key at index %d.", i)
+				}
+
+				state = parserStateBeginValue
+
+			case ';':
+				// Key without value
+				key = normalizeOdbcKey(key)
+				if len(key) == 0 {
+					return res, fmt.Errorf("Unexpected end of key at index %d.", i)
+				}
+
+				res[key] = value
+				key = ""
+				value = ""
+				state = parserStateBeforeKey
+
+			default:
+				key += string(c)
+			}
+
+		case parserStateBeginValue:
+			switch {
+			case c == '{':
+				state = parserStateBracedValue
+			case c == ';':
+				// Empty value
+				res[key] = value
+				key = ""
+				state = parserStateBeforeKey
+			case unicode.IsSpace(c):
+				// Ignore whitespace
+			default:
+				state = parserStateBareValue
+				value += string(c)
+			}
+
+		case parserStateBareValue:
+			if c == ';' {
+				res[key] = strings.TrimRightFunc(value, unicode.IsSpace)
+				key = ""
+				value = ""
+				state = parserStateBeforeKey
+			} else {
+				value += string(c)
+			}
+
+		case parserStateBracedValue:
+			if c == '}' {
+				state = parserStateBracedValueClosingBrace
+			} else {
+				value += string(c)
+			}
+
+		case parserStateBracedValueClosingBrace:
+			if c == '}' {
+				// Escaped closing brace
+				value += string(c)
+				state = parserStateBracedValue
+				continue
+			}
+
+			// End of braced value
+			res[key] = value
+			key = ""
+			value = ""
+
+			// This character is the first character past the end,
+			// so it needs to be parsed like the parserStateEndValue state.
+			state = parserStateEndValue
+			switch {
+			case c == ';':
+				state = parserStateBeforeKey
+			case unicode.IsSpace(c):
+				// Ignore whitespace
+			default:
+				return res, fmt.Errorf("Unexpected character %c at index %d. Expected semi-colon or whitespace.", c, i)
+			}
+
+		case parserStateEndValue:
+			switch {
+			case c == ';':
+				state = parserStateBeforeKey
+			case unicode.IsSpace(c):
+				// Ignore whitespace
+			default:
+				return res, fmt.Errorf("Unexpected character %c at index %d. Expected semi-colon or whitespace.", c, i)
+			}
+		}
+	}
+
+	switch state {
+	case parserStateBeforeKey: // Okay
+	case parserStateKey: // Unfinished key. Treat as key without value.
+		key = normalizeOdbcKey(key)
+		if len(key) == 0 {
+			return res, fmt.Errorf("Unexpected end of key at index %d.", len(dsn))
+		}
+		res[key] = value
+	case parserStateBeginValue: // Empty value
+		res[key] = value
+	case parserStateBareValue:
+		res[key] = strings.TrimRightFunc(value, unicode.IsSpace)
+	case parserStateBracedValue:
+		return res, fmt.Errorf("Unexpected end of braced value at index %d.", len(dsn))
+	case parserStateBracedValueClosingBrace: // End of braced value
+		res[key] = value
+	case parserStateEndValue: // Okay
+	}
+
+	return res, nil
+}
+
+// Normalizes the given string as an ODBC-format key
+func normalizeOdbcKey(s string) string {
+	return strings.ToLower(strings.TrimRightFunc(s, unicode.IsSpace))
+}
+
+// Splits a URL of the form sqlserver://username:password@host/instance?param1=value&param2=value
+func splitConnectionStringURL(dsn string) (map[string]string, error) {
+	res := map[string]string{}
+
+	u, err := url.Parse(dsn)
+	if err != nil {
+		return res, err
+	}
+
+	if u.Scheme != "sqlserver" {
+		return res, fmt.Errorf("scheme %s is not recognized", u.Scheme)
+	}
+
+	if u.User != nil {
+		res["user id"] = u.User.Username()
+		p, exists := u.User.Password()
+		if exists {
+			res["password"] = p
+		}
+	}
+
+	host, port, err := net.SplitHostPort(u.Host)
+	if err != nil {
+		host = u.Host
+	}
+
+	if len(u.Path) > 0 {
+		res["server"] = host + "\\" + u.Path[1:]
+	} else {
+		res["server"] = host
+	}
+
+	if len(port) > 0 {
+		res["port"] = port
+	}
+
+	query := u.Query()
+	for k, v := range query {
+		if len(v) > 1 {
+			return res, fmt.Errorf("key %s provided more than once", k)
+		}
+		res[strings.ToLower(k)] = v[0]
+	}
+
+	return res, nil
+}
+
 func parseConnectParams(dsn string) (connectParams, error) {
-	params := splitConnectionString(dsn)
 	var p connectParams
+
+	var params map[string]string
+	if strings.HasPrefix(dsn, "odbc:") {
+		parameters, err := splitConnectionStringOdbc(dsn[len("odbc:"):])
+		if err != nil {
+			return p, err
+		}
+		params = parameters
+	} else if strings.HasPrefix(dsn, "sqlserver://") {
+		parameters, err := splitConnectionStringURL(dsn)
+		if err != nil {
+			return p, err
+		}
+		params = parameters
+	} else {
+		params = splitConnectionString(dsn)
+	}
+
 	strlog, ok := params["log"]
 	if ok {
 		var err error
@@ -712,7 +967,32 @@ func parseConnectParams(dsn string) (connectParams, error) {
 		}
 	}
 
-	p.dial_timeout = 5 * time.Second
+	// https://docs.microsoft.com/en-us/sql/database-engine/configure-windows/configure-the-network-packet-size-server-configuration-option
+	// Default packet size remains at 4096 bytes
+	p.packetSize = 4096
+	strpsize, ok := params["packet size"]
+	if ok {
+		var err error
+		psize, err := strconv.ParseUint(strpsize, 0, 16)
+		if err != nil {
+			f := "Invalid packet size '%v': %v"
+			return p, fmt.Errorf(f, strpsize, err.Error())
+		}
+
+		// Ensure packet size falls within the TDS protocol range of 512 to 32767 bytes
+		// NOTE: Encrypted connections have a maximum size of 16383 bytes.  If you request
+		// a higher packet size, the server will respond with an ENVCHANGE request to
+		// alter the packet size to 16383 bytes.
+		p.packetSize = uint16(psize)
+		if p.packetSize < 512 {
+			p.packetSize = 512
+		} else if p.packetSize > 32767 {
+			p.packetSize = 32767
+		}
+	}
+
+	// https://msdn.microsoft.com/en-us/library/dd341108.aspx
+	p.dial_timeout = 15 * time.Second
 	p.conn_timeout = 30 * time.Second
 	strconntimeout, ok := params["connection timeout"]
 	if ok {
@@ -732,8 +1012,12 @@ func parseConnectParams(dsn string) (connectParams, error) {
 		}
 		p.dial_timeout = time.Duration(timeout) * time.Second
 	}
-	keepAlive, ok := params["keepalive"]
-	if ok {
+
+	// default keep alive should be 30 seconds according to spec:
+	// https://msdn.microsoft.com/en-us/library/dd341108.aspx
+	p.keepAlive = 30 * time.Second
+
+	if keepAlive, ok := params["keepalive"]; ok {
 		timeout, err := strconv.ParseUint(keepAlive, 0, 16)
 		if err != nil {
 			f := "Invalid keepAlive value '%s': %s"
@@ -743,7 +1027,7 @@ func parseConnectParams(dsn string) (connectParams, error) {
 	}
 	encrypt, ok := params["encrypt"]
 	if ok {
-		if strings.ToUpper(encrypt) == "DISABLE" {
+		if strings.EqualFold(encrypt, "DISABLE") {
 			p.disableEncryption = true
 		} else {
 			var err error
@@ -819,7 +1103,7 @@ func parseConnectParams(dsn string) (connectParams, error) {
 	return p, nil
 }
 
-type Auth interface {
+type auth interface {
 	InitialBytes() ([]byte, error)
 	NextBytes([]byte) ([]byte, error)
 	Free()
@@ -828,7 +1112,7 @@ type Auth interface {
 // SQL Server AlwaysOn Availability Group Listeners are bound by DNS to a
 // list of IP addresses.  So if there is more than one, try them all and
 // use the first one that allows a connection.
-func dialConnection(p connectParams) (conn net.Conn, err error) {
+func dialConnection(ctx context.Context, p connectParams) (conn net.Conn, err error) {
 	var ips []net.IP
 	ips, err = net.LookupIP(p.host)
 	if err != nil {
@@ -839,9 +1123,9 @@ func dialConnection(p connectParams) (conn net.Conn, err error) {
 		ips = []net.IP{ip}
 	}
 	if len(ips) == 1 {
-		d := createDialer(p)
+		d := createDialer(&p)
 		addr := net.JoinHostPort(ips[0].String(), strconv.Itoa(int(p.port)))
-		conn, err = d.Dial("tcp", addr)
+		conn, err = d.Dial(ctx, addr)
 
 	} else {
 		//Try Dials in parallel to avoid waiting for timeouts.
@@ -850,9 +1134,9 @@ func dialConnection(p connectParams) (conn net.Conn, err error) {
 		portStr := strconv.Itoa(int(p.port))
 		for _, ip := range ips {
 			go func(ip net.IP) {
-				d := createDialer(p)
+				d := createDialer(&p)
 				addr := net.JoinHostPort(ip.String(), portStr)
-				conn, err := d.Dial("tcp", addr)
+				conn, err := d.Dial(ctx, addr)
 				if err == nil {
 					connChan <- conn
 				} else {
@@ -887,16 +1171,15 @@ func dialConnection(p connectParams) (conn net.Conn, err error) {
 		f := "Unable to open tcp connection with host '%v:%v': %v"
 		return nil, fmt.Errorf(f, p.host, p.port, err.Error())
 	}
-
 	return conn, err
 }
 
-func connect(p connectParams) (res *tdsSession, err error) {
+func connect(ctx context.Context, log optionalLogger, p connectParams) (res *tdsSession, err error) {
 	res = nil
 	// if instance is specified use instance resolution service
 	if p.instance != "" {
 		p.instance = strings.ToUpper(p.instance)
-		instances, err := getInstances(p.host)
+		instances, err := getInstances(ctx, p.host)
 		if err != nil {
 			f := "Unable to get instances from Sql Server Browser on host %v: %v"
 			return nil, fmt.Errorf(f, p.host, err.Error())
@@ -914,16 +1197,17 @@ func connect(p connectParams) (res *tdsSession, err error) {
 	}
 
 initiate_connection:
-	conn, err := dialConnection(p)
+	conn, err := dialConnection(ctx, p)
 	if err != nil {
 		return nil, err
 	}
 
 	toconn := NewTimeoutConn(conn, p.conn_timeout)
 
-	outbuf := newTdsBuffer(4096, toconn)
+	outbuf := newTdsBuffer(p.packetSize, toconn)
 	sess := tdsSession{
 		buf:      outbuf,
+		log:      log,
 		logFlags: p.logFlags,
 	}
 
@@ -969,8 +1253,7 @@ initiate_connection:
 		if p.certificate != "" {
 			pem, err := ioutil.ReadFile(p.certificate)
 			if err != nil {
-				f := "Cannot read certificate '%s': %s"
-				return nil, fmt.Errorf(f, p.certificate, err.Error())
+				return nil, fmt.Errorf("Cannot read certificate %q: %v", p.certificate, err)
 			}
 			certs := x509.NewCertPool()
 			certs.AppendCertsFromPEM(pem)
@@ -980,15 +1263,20 @@ initiate_connection:
 			config.InsecureSkipVerify = true
 		}
 		config.ServerName = p.hostInCertificate
+		// fix for https://github.com/denisenkom/go-mssqldb/issues/166
+		// Go implementation of TLS payload size heuristic algorithm splits single TDS package to multiple TCP segments,
+		// while SQL Server seems to expect one TCP segment per encrypted TDS package.
+		// Setting DynamicRecordSizingDisabled to true disables that algorithm and uses 16384 bytes per TLS package
+		config.DynamicRecordSizingDisabled = true
 		outbuf.transport = conn
 		toconn.buf = outbuf
 		tlsConn := tls.Client(toconn, &config)
 		err = tlsConn.Handshake()
+
 		toconn.buf = nil
 		outbuf.transport = tlsConn
 		if err != nil {
-			f := "TLS Handshake failed: %s"
-			return nil, fmt.Errorf(f, err.Error())
+			return nil, fmt.Errorf("TLS Handshake failed: %v", err)
 		}
 		if encrypt == encryptOff {
 			outbuf.afterFirst = func() {
@@ -999,7 +1287,7 @@ initiate_connection:
 
 	login := login{
 		TDSVersion:   verTDS74,
-		PacketSize:   uint32(len(outbuf.buf)),
+		PacketSize:   uint32(outbuf.PackageSize()),
 		Database:     p.database,
 		OptionFlags2: fODBC, // to get unlimited TEXTSIZE
 		HostName:     p.workstation,
@@ -1028,7 +1316,7 @@ initiate_connection:
 	var sspi_msg []byte
 continue_login:
 	tokchan := make(chan tokenStruct, 5)
-	go processResponse(&sess, tokchan)
+	go processResponse(context.Background(), &sess, tokchan, nil)
 	success := false
 	for tok := range tokchan {
 		switch token := tok.(type) {
@@ -1042,6 +1330,10 @@ continue_login:
 			sess.loginAck = token
 		case error:
 			return nil, fmt.Errorf("Login error: %s", token.Error())
+		case doneStruct:
+			if token.isError() {
+				return nil, fmt.Errorf("Login error: %s", token.getError())
+			}
 		}
 	}
 	if sspi_msg != nil {
diff --git a/vendor/github.com/denisenkom/go-mssqldb/token.go b/vendor/github.com/denisenkom/go-mssqldb/token.go
index f20bd14cc9..5f2167eb86 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/token.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/token.go
@@ -1,30 +1,40 @@
 package mssql
 
 import (
+	"context"
 	"encoding/binary"
+	"errors"
+	"fmt"
 	"io"
+	"net"
 	"strconv"
 	"strings"
 )
 
+//go:generate stringer -type token
+
+type token byte
+
 // token ids
 const (
-	tokenReturnStatus = 121 // 0x79
-	tokenColMetadata  = 129 // 0x81
-	tokenOrder        = 169 // 0xA9
-	tokenError        = 170 // 0xAA
-	tokenInfo         = 171 // 0xAB
-	tokenLoginAck     = 173 // 0xad
-	tokenRow          = 209 // 0xd1
-	tokenNbcRow       = 210 // 0xd2
-	tokenEnvChange    = 227 // 0xE3
-	tokenSSPI         = 237 // 0xED
-	tokenDone         = 253 // 0xFD
-	tokenDoneProc     = 254
-	tokenDoneInProc   = 255
+	tokenReturnStatus token = 121 // 0x79
+	tokenColMetadata  token = 129 // 0x81
+	tokenOrder        token = 169 // 0xA9
+	tokenError        token = 170 // 0xAA
+	tokenInfo         token = 171 // 0xAB
+	tokenReturnValue  token = 0xAC
+	tokenLoginAck     token = 173 // 0xad
+	tokenRow          token = 209 // 0xd1
+	tokenNbcRow       token = 210 // 0xd2
+	tokenEnvChange    token = 227 // 0xE3
+	tokenSSPI         token = 237 // 0xED
+	tokenDone         token = 253 // 0xFD
+	tokenDoneProc     token = 254
+	tokenDoneInProc   token = 255
 )
 
 // done flags
+// https://msdn.microsoft.com/en-us/library/dd340421.aspx
 const (
 	doneFinal    = 0
 	doneMore     = 1
@@ -59,6 +69,13 @@ const (
 	envRouting               = 20
 )
 
+// COLMETADATA flags
+// https://msdn.microsoft.com/en-us/library/dd357363.aspx
+const (
+	colFlagNullable = 1
+	// TODO implement more flags
+)
+
 // interface for all tokens
 type tokenStruct interface{}
 
@@ -70,6 +87,19 @@ type doneStruct struct {
 	Status   uint16
 	CurCmd   uint16
 	RowCount uint64
+	errors   []Error
+}
+
+func (d doneStruct) isError() bool {
+	return d.Status&doneError != 0 || len(d.errors) > 0
+}
+
+func (d doneStruct) getError() Error {
+	if len(d.errors) > 0 {
+		return d.errors[len(d.errors)-1]
+	} else {
+		return Error{Message: "Request failed but didn't provide reason"}
+	}
 }
 
 type doneInProcStruct doneStruct
@@ -120,27 +150,23 @@ func processEnvChg(sess *tdsSession) {
 				badStreamPanic(err)
 			}
 		case envTypLanguage:
-			//currently ignored
-			// old value
-			_, err = readBVarChar(r)
-			if err != nil {
+			// currently ignored
+			// new value
+			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
-			// new value
-			_, err = readBVarChar(r)
-			if err != nil {
+			// old value
+			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
 		case envTypCharset:
-			//currently ignored
-			// old value
-			_, err = readBVarChar(r)
-			if err != nil {
+			// currently ignored
+			// new value
+			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
-			// new value
-			_, err = readBVarChar(r)
-			if err != nil {
+			// old value
+			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
 		case envTypPacketSize:
@@ -156,38 +182,55 @@ func processEnvChg(sess *tdsSession) {
 			if err != nil {
 				badStreamPanicf("Invalid Packet size value returned from server (%s): %s", packetsize, err.Error())
 			}
-			if len(sess.buf.buf) != packetsizei {
-				newbuf := make([]byte, packetsizei)
-				copy(newbuf, sess.buf.buf)
-				sess.buf.buf = newbuf
-			}
+			sess.buf.ResizeBuffer(packetsizei)
 		case envSortId:
 			// currently ignored
-			// old value, should be 0
+			// new value
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
-			// new value
+			// old value, should be 0
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
 		case envSortFlags:
 			// currently ignored
-			// old value, should be 0
+			// new value
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
-			// new value
+			// old value, should be 0
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
 		case envSqlCollation:
 			// currently ignored
-			// old value
-			if _, err = readBVarChar(r); err != nil {
+			var collationSize uint8
+			err = binary.Read(r, binary.LittleEndian, &collationSize)
+			if err != nil {
 				badStreamPanic(err)
 			}
-			// new value
+
+			// SQL Collation data should contain 5 bytes in length
+			if collationSize != 5 {
+				badStreamPanicf("Invalid SQL Collation size value returned from server: %s", collationSize)
+			}
+
+			// 4 bytes, contains: LCID ColFlags Version
+			var info uint32
+			err = binary.Read(r, binary.LittleEndian, &info)
+			if err != nil {
+				badStreamPanic(err)
+			}
+
+			// 1 byte, contains: sortID
+			var sortID uint8
+			err = binary.Read(r, binary.LittleEndian, &sortID)
+			if err != nil {
+				badStreamPanic(err)
+			}
+
+			// old value, should be 0
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
@@ -226,21 +269,21 @@ func processEnvChg(sess *tdsSession) {
 			sess.tranid = 0
 		case envEnlistDTC:
 			// currently ignored
-			// old value
+			// new value, should be 0
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
-			// new value, should be 0
+			// old value
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
 		case envDefectTran:
 			// currently ignored
-			// old value, should be 0
+			// new value
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
-			// new value
+			// old value, should be 0
 			if _, err = readBVarChar(r); err != nil {
 				badStreamPanic(err)
 			}
@@ -358,6 +401,7 @@ func parseOrder(r *tdsBuffer) (res orderStruct) {
 	return res
 }
 
+// https://msdn.microsoft.com/en-us/library/dd340421.aspx
 func parseDone(r *tdsBuffer) (res doneStruct) {
 	res.Status = r.uint16()
 	res.CurCmd = r.uint16()
@@ -365,6 +409,7 @@ func parseDone(r *tdsBuffer) (res doneStruct) {
 	return res
 }
 
+// https://msdn.microsoft.com/en-us/library/dd340553.aspx
 func parseDoneInProc(r *tdsBuffer) (res doneInProcStruct) {
 	res.Status = r.uint16()
 	res.CurCmd = r.uint16()
@@ -473,26 +518,57 @@ func parseInfo(r *tdsBuffer) (res Error) {
 	return
 }
 
-func processResponse(sess *tdsSession, ch chan tokenStruct) {
+// https://msdn.microsoft.com/en-us/library/dd303881.aspx
+func parseReturnValue(r *tdsBuffer) (nv namedValue) {
+	/*
+		ParamOrdinal
+		ParamName
+		Status
+		UserType
+		Flags
+		TypeInfo
+		CryptoMetadata
+		Value
+	*/
+	r.uint16()
+	nv.Name = r.BVarChar()
+	r.byte()
+	r.uint32() // UserType (uint16 prior to 7.2)
+	r.uint16()
+	ti := readTypeInfo(r)
+	nv.Value = ti.Reader(&ti, r)
+	return
+}
+
+func processSingleResponse(sess *tdsSession, ch chan tokenStruct, outs map[string]interface{}) {
 	defer func() {
 		if err := recover(); err != nil {
+			if sess.logFlags&logErrors != 0 {
+				sess.log.Printf("ERROR: Intercepted panic %v", err)
+			}
 			ch <- err
 		}
 		close(ch)
 	}()
+
 	packet_type, err := sess.buf.BeginRead()
 	if err != nil {
+		if sess.logFlags&logErrors != 0 {
+			sess.log.Printf("ERROR: BeginRead failed %v", err)
+		}
 		ch <- err
 		return
 	}
 	if packet_type != packReply {
-		badStreamPanicf("invalid response packet type, expected REPLY, actual: %d", packet_type)
+		badStreamPanic(fmt.Errorf("unexpected packet type in reply: got %v, expected %v", packet_type, packReply))
 	}
 	var columns []columnStruct
-	var lastError Error
-	var failed bool
+	errs := make([]Error, 0, 5)
 	for {
-		token := sess.buf.byte()
+		token := token(sess.buf.byte())
+		if sess.logFlags&logDebug != 0 {
+			sess.log.Printf("got token %v", token)
+		}
 		switch token {
 		case tokenSSPI:
 			ch <- parseSSPIMsg(sess.buf)
@@ -514,18 +590,17 @@ func processResponse(sess *tdsSession, ch chan tokenStruct) {
 			ch <- done
 		case tokenDone, tokenDoneProc:
 			done := parseDone(sess.buf)
-			if sess.logFlags&logRows != 0 && done.Status&doneCount != 0 {
-				sess.log.Printf("(%d row(s) affected)\n", done.RowCount)
-			}
-			if done.Status&doneError != 0 || failed {
-				ch <- lastError
-				return
+			done.errors = errs
+			if sess.logFlags&logDebug != 0 {
+				sess.log.Printf("got DONE or DONEPROC status=%d", done.Status)
 			}
 			if done.Status&doneSrvError != 0 {
-				lastError.Message = "Server Error"
-				ch <- lastError
+				ch <- errors.New("SQL Server had internal error")
 				return
 			}
+			if sess.logFlags&logRows != 0 && done.Status&doneCount != 0 {
+				sess.log.Printf("(%d row(s) affected)\n", done.RowCount)
+			}
 			ch <- done
 			if done.Status&doneMore == 0 {
 				return
@@ -544,18 +619,210 @@ func processResponse(sess *tdsSession, ch chan tokenStruct) {
 		case tokenEnvChange:
 			processEnvChg(sess)
 		case tokenError:
-			lastError = parseError72(sess.buf)
-			failed = true
+			err := parseError72(sess.buf)
+			if sess.logFlags&logDebug != 0 {
+				sess.log.Printf("got ERROR %d %s", err.Number, err.Message)
+			}
+			errs = append(errs, err)
 			if sess.logFlags&logErrors != 0 {
-				sess.log.Println(lastError.Message)
+				sess.log.Println(err.Message)
 			}
 		case tokenInfo:
 			info := parseInfo(sess.buf)
+			if sess.logFlags&logDebug != 0 {
+				sess.log.Printf("got INFO %d %s", info.Number, info.Message)
+			}
 			if sess.logFlags&logMessages != 0 {
 				sess.log.Println(info.Message)
 			}
+		case tokenReturnValue:
+			nv := parseReturnValue(sess.buf)
+			if len(nv.Name) > 0 {
+				name := nv.Name[1:] // Remove the leading "@".
+				if ov, has := outs[name]; has {
+					err = scanIntoOut(nv.Value, ov)
+					if err != nil {
+						fmt.Println("scan error", err)
+						ch <- err
+					}
+				}
+			}
 		default:
-			badStreamPanicf("Unknown token type: %d", token)
+			badStreamPanic(fmt.Errorf("unknown token type returned: %v", token))
+		}
+	}
+}
+
+func scanIntoOut(fromServer, scanInto interface{}) error {
+	switch fs := fromServer.(type) {
+	case int64:
+		switch si := scanInto.(type) {
+		case *int64:
+			*si = fs
+		default:
+			return fmt.Errorf("unsupported scan into type %[1]T for server type %[2]T", scanInto, fromServer)
+		}
+		return nil
+	case string:
+		switch si := scanInto.(type) {
+		case *string:
+			*si = fs
+		default:
+			return fmt.Errorf("unsupported scan into type %[1]T for server type %[2]T", scanInto, fromServer)
+		}
+		return nil
+	}
+	return fmt.Errorf("unsupported type from server %[1]T=%[1]v", fromServer)
+}
+
+type parseRespIter byte
+
+const (
+	parseRespIterContinue parseRespIter = iota // Continue parsing current token.
+	parseRespIterNext                          // Fetch the next token.
+	parseRespIterDone                          // Done with parsing the response.
+)
+
+type parseRespState byte
+
+const (
+	parseRespStateNormal  parseRespState = iota // Normal response state.
+	parseRespStateCancel                        // Query is canceled, wait for server to confirm.
+	parseRespStateClosing                       // Waiting for tokens to come through.
+)
+
+type parseResp struct {
+	sess        *tdsSession
+	ctxDone     <-chan struct{}
+	state       parseRespState
+	cancelError error
+}
+
+func (ts *parseResp) sendAttention(ch chan tokenStruct) parseRespIter {
+	if err := sendAttention(ts.sess.buf); err != nil {
+		ts.dlogf("failed to send attention signal %v", err)
+		ch <- err
+		return parseRespIterDone
+	}
+	ts.state = parseRespStateCancel
+	return parseRespIterContinue
+}
+
+func (ts *parseResp) dlog(msg string) {
+	if ts.sess.logFlags&logDebug != 0 {
+		ts.sess.log.Println(msg)
+	}
+}
+func (ts *parseResp) dlogf(f string, v ...interface{}) {
+	if ts.sess.logFlags&logDebug != 0 {
+		ts.sess.log.Printf(f, v...)
+	}
+}
+
+func (ts *parseResp) iter(ctx context.Context, ch chan tokenStruct, tokChan chan tokenStruct) parseRespIter {
+	switch ts.state {
+	default:
+		panic("unknown state")
+	case parseRespStateNormal:
+		select {
+		case tok, ok := <-tokChan:
+			if !ok {
+				ts.dlog("response finished")
+				return parseRespIterDone
+			}
+			if err, ok := tok.(net.Error); ok && err.Timeout() {
+				ts.cancelError = err
+				ts.dlog("got timeout error, sending attention signal to server")
+				return ts.sendAttention(ch)
+			}
+			// Pass the token along.
+			ch <- tok
+			return parseRespIterContinue
+
+		case <-ts.ctxDone:
+			ts.ctxDone = nil
+			ts.dlog("got cancel message, sending attention signal to server")
+			return ts.sendAttention(ch)
+		}
+	case parseRespStateCancel: // Read all responses until a DONE or error is received.Auth
+		select {
+		case tok, ok := <-tokChan:
+			if !ok {
+				ts.dlog("response finished but waiting for attention ack")
+				return parseRespIterNext
+			}
+			switch tok := tok.(type) {
+			default:
+				// Ignore all other tokens while waiting.
+				// The TDS spec says other tokens may arrive after an attention
+				// signal is sent. Ignore these tokens and continue looking for
+				// a DONE with attention confirm mark.
+			case doneStruct:
+				if tok.Status&doneAttn != 0 {
+					ts.dlog("got cancellation confirmation from server")
+					if ts.cancelError != nil {
+						ch <- ts.cancelError
+						ts.cancelError = nil
+					} else {
+						ch <- ctx.Err()
+					}
+					return parseRespIterDone
+				}
+
+			// If an error happens during cancel, pass it along and just stop.
+			// We are uncertain to receive more tokens.
+			case error:
+				ch <- tok
+				ts.state = parseRespStateClosing
+			}
+			return parseRespIterContinue
+		case <-ts.ctxDone:
+			ts.ctxDone = nil
+			ts.state = parseRespStateClosing
+			return parseRespIterContinue
+		}
+	case parseRespStateClosing: // Wait for current token chan to close.
+		if _, ok := <-tokChan; !ok {
+			ts.dlog("response finished")
+			return parseRespIterDone
+		}
+		return parseRespIterContinue
+	}
+}
+
+func processResponse(ctx context.Context, sess *tdsSession, ch chan tokenStruct, outs map[string]interface{}) {
+	ts := &parseResp{
+		sess:    sess,
+		ctxDone: ctx.Done(),
+	}
+	defer func() {
+		// Ensure any remaining error is piped through
+		// or the query may look like it executed when it actually failed.
+		if ts.cancelError != nil {
+			ch <- ts.cancelError
+			ts.cancelError = nil
+		}
+		close(ch)
+	}()
+
+	// Loop over multiple responses.
+	for {
+		ts.dlog("initiating response reading")
+
+		tokChan := make(chan tokenStruct)
+		go processSingleResponse(sess, tokChan, outs)
+
+		// Loop over multiple tokens in response.
+	tokensLoop:
+		for {
+			switch ts.iter(ctx, ch, tokChan) {
+			case parseRespIterContinue:
+				// Nothing, continue to next token.
+			case parseRespIterNext:
+				break tokensLoop
+			case parseRespIterDone:
+				return
+			}
 		}
 	}
 }
diff --git a/vendor/github.com/denisenkom/go-mssqldb/token_string.go b/vendor/github.com/denisenkom/go-mssqldb/token_string.go
new file mode 100644
index 0000000000..c075b23be0
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/token_string.go
@@ -0,0 +1,53 @@
+// Code generated by "stringer -type token"; DO NOT EDIT
+
+package mssql
+
+import "fmt"
+
+const (
+	_token_name_0 = "tokenReturnStatus"
+	_token_name_1 = "tokenColMetadata"
+	_token_name_2 = "tokenOrdertokenErrortokenInfo"
+	_token_name_3 = "tokenLoginAck"
+	_token_name_4 = "tokenRowtokenNbcRow"
+	_token_name_5 = "tokenEnvChange"
+	_token_name_6 = "tokenSSPI"
+	_token_name_7 = "tokenDonetokenDoneProctokenDoneInProc"
+)
+
+var (
+	_token_index_0 = [...]uint8{0, 17}
+	_token_index_1 = [...]uint8{0, 16}
+	_token_index_2 = [...]uint8{0, 10, 20, 29}
+	_token_index_3 = [...]uint8{0, 13}
+	_token_index_4 = [...]uint8{0, 8, 19}
+	_token_index_5 = [...]uint8{0, 14}
+	_token_index_6 = [...]uint8{0, 9}
+	_token_index_7 = [...]uint8{0, 9, 22, 37}
+)
+
+func (i token) String() string {
+	switch {
+	case i == 121:
+		return _token_name_0
+	case i == 129:
+		return _token_name_1
+	case 169 <= i && i <= 171:
+		i -= 169
+		return _token_name_2[_token_index_2[i]:_token_index_2[i+1]]
+	case i == 173:
+		return _token_name_3
+	case 209 <= i && i <= 210:
+		i -= 209
+		return _token_name_4[_token_index_4[i]:_token_index_4[i+1]]
+	case i == 227:
+		return _token_name_5
+	case i == 237:
+		return _token_name_6
+	case 253 <= i && i <= 255:
+		i -= 253
+		return _token_name_7[_token_index_7[i]:_token_index_7[i+1]]
+	default:
+		return fmt.Sprintf("token(%d)", i)
+	}
+}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/tran.go b/vendor/github.com/denisenkom/go-mssqldb/tran.go
index ae38107661..75e7a2ae65 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/tran.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/tran.go
@@ -1,6 +1,7 @@
+package mssql
+
 // Transaction Manager requests
 // http://msdn.microsoft.com/en-us/library/dd339887.aspx
-package mssql
 
 import (
 	"encoding/binary"
@@ -16,7 +17,18 @@ const (
 	tmSaveXact      = 9
 )
 
-func sendBeginXact(buf *tdsBuffer, headers []headerStruct, isolation uint8,
+type isoLevel uint8
+
+const (
+	isolationUseCurrent     isoLevel = 0
+	isolationReadUncommited          = 1
+	isolationReadCommited            = 2
+	isolationRepeatableRead          = 3
+	isolationSerializable            = 4
+	isolationSnapshot                = 5
+)
+
+func sendBeginXact(buf *tdsBuffer, headers []headerStruct, isolation isoLevel,
 	name string) (err error) {
 	buf.BeginPacket(packTransMgrReq)
 	writeAllHeaders(buf, headers)
diff --git a/vendor/github.com/denisenkom/go-mssqldb/types.go b/vendor/github.com/denisenkom/go-mssqldb/types.go
index c38862e9eb..9ae5c5f761 100644
--- a/vendor/github.com/denisenkom/go-mssqldb/types.go
+++ b/vendor/github.com/denisenkom/go-mssqldb/types.go
@@ -6,8 +6,11 @@ import (
 	"fmt"
 	"io"
 	"math"
+	"reflect"
 	"strconv"
 	"time"
+
+	"github.com/denisenkom/go-mssqldb/internal/cp"
 )
 
 // fixed-length data types
@@ -66,6 +69,9 @@ const (
 	typeNText   = 0x63
 	typeVariant = 0x62
 )
+const PLP_NULL = 0xFFFFFFFFFFFFFFFF
+const UNKNOWN_PLP_LEN = 0xFFFFFFFFFFFFFFFE
+const PLP_TERMINATOR = 0x00000000
 
 // TYPE_INFO rule
 // http://msdn.microsoft.com/en-us/library/dd358284.aspx
@@ -75,11 +81,32 @@ type typeInfo struct {
 	Scale     uint8
 	Prec      uint8
 	Buffer    []byte
-	Collation collation
+	Collation cp.Collation
+	UdtInfo   udtInfo
+	XmlInfo   xmlInfo
 	Reader    func(ti *typeInfo, r *tdsBuffer) (res interface{})
 	Writer    func(w io.Writer, ti typeInfo, buf []byte) (err error)
 }
 
+// Common Language Runtime (CLR) Instances
+// http://msdn.microsoft.com/en-us/library/dd357962.aspx
+type udtInfo struct {
+	//MaxByteSize         uint32
+	DBName                string
+	SchemaName            string
+	TypeName              string
+	AssemblyQualifiedName string
+}
+
+// XML Values
+// http://msdn.microsoft.com/en-us/library/dd304764.aspx
+type xmlInfo struct {
+	SchemaPresent       uint8
+	DBName              string
+	OwningSchema        string
+	XmlSchemaCollection string
+}
+
 func readTypeInfo(r *tdsBuffer) (res typeInfo) {
 	res.TypeId = r.byte()
 	switch res.TypeId {
@@ -114,7 +141,8 @@ func writeTypeInfo(w io.Writer, ti *typeInfo) (err error) {
 	switch ti.TypeId {
 	case typeNull, typeInt1, typeBit, typeInt2, typeInt4, typeDateTim4,
 		typeFlt4, typeMoney, typeDateTime, typeFlt8, typeMoney4, typeInt8:
-		// those are fixed length types
+		// those are fixed length
+		ti.Writer = writeFixedType
 	default: // all others are VARLENTYPE
 		err = writeVarLen(w, ti)
 		if err != nil {
@@ -124,19 +152,25 @@ func writeTypeInfo(w io.Writer, ti *typeInfo) (err error) {
 	return
 }
 
+func writeFixedType(w io.Writer, ti typeInfo, buf []byte) (err error) {
+	_, err = w.Write(buf)
+	return
+}
+
 func writeVarLen(w io.Writer, ti *typeInfo) (err error) {
 	switch ti.TypeId {
 	case typeDateN:
-
+		ti.Writer = writeByteLenType
 	case typeTimeN, typeDateTime2N, typeDateTimeOffsetN:
 		if err = binary.Write(w, binary.LittleEndian, ti.Scale); err != nil {
 			return
 		}
 		ti.Writer = writeByteLenType
-	case typeGuid, typeIntN, typeDecimal, typeNumeric,
+	case typeIntN, typeDecimal, typeNumeric,
 		typeBitN, typeDecimalN, typeNumericN, typeFltN,
 		typeMoneyN, typeDateTimeN, typeChar,
 		typeVarChar, typeBinary, typeVarBinary:
+
 		// byle len types
 		if ti.Size > 0xff {
 			panic("Invalid size for BYLELEN_TYPE")
@@ -156,6 +190,14 @@ func writeVarLen(w io.Writer, ti *typeInfo) (err error) {
 			}
 		}
 		ti.Writer = writeByteLenType
+	case typeGuid:
+		if !(ti.Size == 0x10 || ti.Size == 0x00) {
+			panic("Invalid size for BYLELEN_TYPE")
+		}
+		if err = binary.Write(w, binary.LittleEndian, uint8(ti.Size)); err != nil {
+			return
+		}
+		ti.Writer = writeByteLenType
 	case typeBigVarBin, typeBigVarChar, typeBigBinary, typeBigChar,
 		typeNVarChar, typeNChar, typeXml, typeUdt:
 		// short len types
@@ -176,14 +218,19 @@ func writeVarLen(w io.Writer, ti *typeInfo) (err error) {
 				return
 			}
 		case typeXml:
-			var schemapresent uint8 = 0
-			if err = binary.Write(w, binary.LittleEndian, schemapresent); err != nil {
+			if err = binary.Write(w, binary.LittleEndian, ti.XmlInfo.SchemaPresent); err != nil {
 				return
 			}
 		}
 	case typeText, typeImage, typeNText, typeVariant:
 		// LONGLEN_TYPE
-		panic("LONGLEN_TYPE not implemented")
+		if err = binary.Write(w, binary.LittleEndian, uint32(ti.Size)); err != nil {
+			return
+		}
+		if err = writeCollation(w, ti.Collation); err != nil {
+			return
+		}
+		ti.Writer = writeLongLenType
 	default:
 		panic("Invalid type")
 	}
@@ -207,7 +254,7 @@ func decodeDateTime(buf []byte) time.Time {
 		0, 0, secs, ns, time.UTC)
 }
 
-func readFixedType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
+func readFixedType(ti *typeInfo, r *tdsBuffer) interface{} {
 	r.ReadFull(ti.Buffer)
 	buf := ti.Buffer
 	switch ti.TypeId {
@@ -241,12 +288,7 @@ func readFixedType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
 	panic("shoulnd't get here")
 }
 
-func writeFixedType(w io.Writer, ti typeInfo, buf []byte) (err error) {
-	_, err = w.Write(buf)
-	return
-}
-
-func readByteLenType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
+func readByteLenType(ti *typeInfo, r *tdsBuffer) interface{} {
 	size := r.byte()
 	if size == 0 {
 		return nil
@@ -305,6 +347,10 @@ func readByteLenType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
 		default:
 			badStreamPanicf("Invalid size for MONEYNTYPE")
 		}
+	case typeDateTim4:
+		return decodeDateTim4(buf)
+	case typeDateTime:
+		return decodeDateTime(buf)
 	case typeDateTimeN:
 		switch len(buf) {
 		case 4:
@@ -341,7 +387,7 @@ func writeByteLenType(w io.Writer, ti typeInfo, buf []byte) (err error) {
 	return
 }
 
-func readShortLenType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
+func readShortLenType(ti *typeInfo, r *tdsBuffer) interface{} {
 	size := r.uint16()
 	if size == 0xffff {
 		return nil
@@ -384,7 +430,7 @@ func writeShortLenType(w io.Writer, ti typeInfo, buf []byte) (err error) {
 	return
 }
 
-func readLongLenType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
+func readLongLenType(ti *typeInfo, r *tdsBuffer) interface{} {
 	// information about this format can be found here:
 	// http://msdn.microsoft.com/en-us/library/dd304783.aspx
 	// and here:
@@ -415,10 +461,51 @@ func readLongLenType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
 	}
 	panic("shoulnd't get here")
 }
+func writeLongLenType(w io.Writer, ti typeInfo, buf []byte) (err error) {
+	//textptr
+	err = binary.Write(w, binary.LittleEndian, byte(0x10))
+	if err != nil {
+		return
+	}
+	err = binary.Write(w, binary.LittleEndian, uint64(0xFFFFFFFFFFFFFFFF))
+	if err != nil {
+		return
+	}
+	err = binary.Write(w, binary.LittleEndian, uint64(0xFFFFFFFFFFFFFFFF))
+	if err != nil {
+		return
+	}
+	//timestamp?
+	err = binary.Write(w, binary.LittleEndian, uint64(0xFFFFFFFFFFFFFFFF))
+	if err != nil {
+		return
+	}
+
+	err = binary.Write(w, binary.LittleEndian, uint32(ti.Size))
+	if err != nil {
+		return
+	}
+	_, err = w.Write(buf)
+	return
+}
+
+func readCollation(r *tdsBuffer) (res cp.Collation) {
+	res.LcidAndFlags = r.uint32()
+	res.SortId = r.byte()
+	return
+}
+
+func writeCollation(w io.Writer, col cp.Collation) (err error) {
+	if err = binary.Write(w, binary.LittleEndian, col.LcidAndFlags); err != nil {
+		return
+	}
+	err = binary.Write(w, binary.LittleEndian, col.SortId)
+	return
+}
 
 // reads variant value
 // http://msdn.microsoft.com/en-us/library/dd303302.aspx
-func readVariantType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
+func readVariantType(ti *typeInfo, r *tdsBuffer) interface{} {
 	size := r.int32()
 	if size == 0 {
 		return nil
@@ -510,14 +597,14 @@ func readVariantType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
 
 // partially length prefixed stream
 // http://msdn.microsoft.com/en-us/library/dd340469.aspx
-func readPLPType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
+func readPLPType(ti *typeInfo, r *tdsBuffer) interface{} {
 	size := r.uint64()
 	var buf *bytes.Buffer
 	switch size {
-	case 0xffffffffffffffff:
+	case PLP_NULL:
 		// null
 		return nil
-	case 0xfffffffffffffffe:
+	case UNKNOWN_PLP_LEN:
 		// size unknown
 		buf = bytes.NewBuffer(make([]byte, 0, 1000))
 	default:
@@ -548,15 +635,16 @@ func readPLPType(ti *typeInfo, r *tdsBuffer) (res interface{}) {
 }
 
 func writePLPType(w io.Writer, ti typeInfo, buf []byte) (err error) {
-	if err = binary.Write(w, binary.LittleEndian, uint64(len(buf))); err != nil {
+	if err = binary.Write(w, binary.LittleEndian, uint64(UNKNOWN_PLP_LEN)); err != nil {
 		return
 	}
 	for {
 		chunksize := uint32(len(buf))
-		if err = binary.Write(w, binary.LittleEndian, chunksize); err != nil {
+		if chunksize == 0 {
+			err = binary.Write(w, binary.LittleEndian, uint32(PLP_TERMINATOR))
 			return
 		}
-		if chunksize == 0 {
+		if err = binary.Write(w, binary.LittleEndian, chunksize); err != nil {
 			return
 		}
 		if _, err = w.Write(buf[:chunksize]); err != nil {
@@ -606,19 +694,27 @@ func readVarLen(ti *typeInfo, r *tdsBuffer) {
 		}
 		ti.Reader = readByteLenType
 	case typeXml:
-		schemapresent := r.byte()
-		if schemapresent != 0 {
-			// just ignore this for now
+		ti.XmlInfo.SchemaPresent = r.byte()
+		if ti.XmlInfo.SchemaPresent != 0 {
 			// dbname
-			r.BVarChar()
+			ti.XmlInfo.DBName = r.BVarChar()
 			// owning schema
-			r.BVarChar()
+			ti.XmlInfo.OwningSchema = r.BVarChar()
 			// xml schema collection
-			r.UsVarChar()
+			ti.XmlInfo.XmlSchemaCollection = r.UsVarChar()
 		}
 		ti.Reader = readPLPType
+	case typeUdt:
+		ti.Size = int(r.uint16())
+		ti.UdtInfo.DBName = r.BVarChar()
+		ti.UdtInfo.SchemaName = r.BVarChar()
+		ti.UdtInfo.TypeName = r.BVarChar()
+		ti.UdtInfo.AssemblyQualifiedName = r.UsVarChar()
+
+		ti.Buffer = make([]byte, ti.Size)
+		ti.Reader = readPLPType
 	case typeBigVarBin, typeBigVarChar, typeBigBinary, typeBigChar,
-		typeNVarChar, typeNChar, typeUdt:
+		typeNVarChar, typeNChar:
 		// short len types
 		ti.Size = int(r.uint16())
 		switch ti.TypeId {
@@ -701,7 +797,8 @@ func decodeDecimal(prec uint8, scale uint8, buf []byte) []byte {
 
 // http://msdn.microsoft.com/en-us/library/ee780895.aspx
 func decodeDateInt(buf []byte) (days int) {
-	return int(buf[0]) + int(buf[1])*256 + int(buf[2])*256*256
+	days = int(buf[0]) + int(buf[1])*256 + int(buf[2])*256*256
+	return
 }
 
 func decodeDate(buf []byte) time.Time {
@@ -767,8 +864,8 @@ func dateTime2(t time.Time) (days int32, ns int64) {
 	return
 }
 
-func decodeChar(col collation, buf []byte) string {
-	return charset2utf8(col, buf)
+func decodeChar(col cp.Collation, buf []byte) string {
+	return cp.CharsetToUTF8(col, buf)
 }
 
 func decodeUcs2(buf []byte) string {
@@ -787,12 +884,127 @@ func decodeXml(ti typeInfo, buf []byte) string {
 	return decodeUcs2(buf)
 }
 
-func decodeUdt(ti typeInfo, buf []byte) int {
-	panic("Not implemented")
+func decodeUdt(ti typeInfo, buf []byte) []byte {
+	return buf
+}
+
+// makes go/sql type instance as described below
+// It should return
+// the value type that can be used to scan types into. For example, the database
+// column type "bigint" this should return "reflect.TypeOf(int64(0))".
+func makeGoLangScanType(ti typeInfo) reflect.Type {
+	switch ti.TypeId {
+	case typeInt1:
+		return reflect.TypeOf(int64(0))
+	case typeInt2:
+		return reflect.TypeOf(int64(0))
+	case typeInt4:
+		return reflect.TypeOf(int64(0))
+	case typeInt8:
+		return reflect.TypeOf(int64(0))
+	case typeFlt4:
+		return reflect.TypeOf(float64(0))
+	case typeIntN:
+		switch ti.Size {
+		case 1:
+			return reflect.TypeOf(int64(0))
+		case 2:
+			return reflect.TypeOf(int64(0))
+		case 4:
+			return reflect.TypeOf(int64(0))
+		case 8:
+			return reflect.TypeOf(int64(0))
+		default:
+			panic("invalid size of INTNTYPE")
+		}
+	case typeFlt8:
+		return reflect.TypeOf(float64(0))
+	case typeFltN:
+		switch ti.Size {
+		case 4:
+			return reflect.TypeOf(float64(0))
+		case 8:
+			return reflect.TypeOf(float64(0))
+		default:
+			panic("invalid size of FLNNTYPE")
+		}
+	case typeBigVarBin:
+		return reflect.TypeOf([]byte{})
+	case typeVarChar:
+		return reflect.TypeOf("")
+	case typeNVarChar:
+		return reflect.TypeOf("")
+	case typeBit, typeBitN:
+		return reflect.TypeOf(true)
+	case typeDecimalN, typeNumericN:
+		return reflect.TypeOf([]byte{})
+	case typeMoney, typeMoney4, typeMoneyN:
+		switch ti.Size {
+		case 4:
+			return reflect.TypeOf([]byte{})
+		case 8:
+			return reflect.TypeOf([]byte{})
+		default:
+			panic("invalid size of MONEYN")
+		}
+	case typeDateTim4:
+		return reflect.TypeOf(time.Time{})
+	case typeDateTime:
+		return reflect.TypeOf(time.Time{})
+	case typeDateTimeN:
+		switch ti.Size {
+		case 4:
+			return reflect.TypeOf(time.Time{})
+		case 8:
+			return reflect.TypeOf(time.Time{})
+		default:
+			panic("invalid size of DATETIMEN")
+		}
+	case typeDateTime2N:
+		return reflect.TypeOf(time.Time{})
+	case typeDateN:
+		return reflect.TypeOf(time.Time{})
+	case typeTimeN:
+		return reflect.TypeOf(time.Time{})
+	case typeDateTimeOffsetN:
+		return reflect.TypeOf(time.Time{})
+	case typeBigVarChar:
+		return reflect.TypeOf("")
+	case typeBigChar:
+		return reflect.TypeOf("")
+	case typeNChar:
+		return reflect.TypeOf("")
+	case typeGuid:
+		return reflect.TypeOf([]byte{})
+	case typeXml:
+		return reflect.TypeOf("")
+	case typeText:
+		return reflect.TypeOf("")
+	case typeNText:
+		return reflect.TypeOf("")
+	case typeImage:
+		return reflect.TypeOf([]byte{})
+	case typeBigBinary:
+		return reflect.TypeOf([]byte{})
+	case typeVariant:
+		return reflect.TypeOf(nil)
+	default:
+		panic(fmt.Sprintf("not implemented makeDecl for type %d", ti.TypeId))
+	}
 }
 
 func makeDecl(ti typeInfo) string {
 	switch ti.TypeId {
+	case typeNull:
+		// maybe we should use something else here
+		// this is tested in TestNull
+		return "nvarchar(1)"
+	case typeInt1:
+		return "tinyint"
+	case typeInt2:
+		return "smallint"
+	case typeInt4:
+		return "int"
 	case typeInt8:
 		return "bigint"
 	case typeFlt4:
@@ -821,24 +1033,415 @@ func makeDecl(ti typeInfo) string {
 		default:
 			panic("invalid size of FLNNTYPE")
 		}
+	case typeDecimal, typeDecimalN:
+		return fmt.Sprintf("decimal(%d, %d)", ti.Prec, ti.Scale)
+	case typeNumeric, typeNumericN:
+		return fmt.Sprintf("numeric(%d, %d)", ti.Prec, ti.Scale)
+	case typeMoney4:
+		return "smallmoney"
+	case typeMoney:
+		return "money"
+	case typeMoneyN:
+		switch ti.Size {
+		case 4:
+			return "smallmoney"
+		case 8:
+			return "money"
+		default:
+			panic("invalid size of MONEYNTYPE")
+		}
 	case typeBigVarBin:
 		if ti.Size > 8000 || ti.Size == 0 {
-			return fmt.Sprintf("varbinary(max)")
+			return "varbinary(max)"
 		} else {
 			return fmt.Sprintf("varbinary(%d)", ti.Size)
 		}
+	case typeNChar:
+		return fmt.Sprintf("nchar(%d)", ti.Size/2)
+	case typeBigChar, typeChar:
+		return fmt.Sprintf("char(%d)", ti.Size)
+	case typeBigVarChar, typeVarChar:
+		if ti.Size > 4000 || ti.Size == 0 {
+			return fmt.Sprintf("varchar(max)")
+		} else {
+			return fmt.Sprintf("varchar(%d)", ti.Size)
+		}
 	case typeNVarChar:
 		if ti.Size > 8000 || ti.Size == 0 {
-			return fmt.Sprintf("nvarchar(max)")
+			return "nvarchar(max)"
 		} else {
 			return fmt.Sprintf("nvarchar(%d)", ti.Size/2)
 		}
 	case typeBit, typeBitN:
 		return "bit"
-	case typeDateTimeN:
+	case typeDateN:
+		return "date"
+	case typeDateTim4:
+		return "smalldatetime"
+	case typeDateTime:
 		return "datetime"
+	case typeDateTimeN:
+		switch ti.Size {
+		case 4:
+			return "smalldatetime"
+		case 8:
+			return "datetime"
+		default:
+			panic("invalid size of DATETIMNTYPE")
+		}
+	case typeDateTime2N:
+		return fmt.Sprintf("datetime2(%d)", ti.Scale)
 	case typeDateTimeOffsetN:
 		return fmt.Sprintf("datetimeoffset(%d)", ti.Scale)
+	case typeText:
+		return "text"
+	case typeNText:
+		return "ntext"
+	case typeUdt:
+		return ti.UdtInfo.TypeName
+	case typeGuid:
+		return "uniqueidentifier"
+	default:
+		panic(fmt.Sprintf("not implemented makeDecl for type %#x", ti.TypeId))
+	}
+}
+
+// makes go/sql type name as described below
+// RowsColumnTypeDatabaseTypeName may be implemented by Rows. It should return the
+// database system type name without the length. Type names should be uppercase.
+// Examples of returned types: "VARCHAR", "NVARCHAR", "VARCHAR2", "CHAR", "TEXT",
+// "DECIMAL", "SMALLINT", "INT", "BIGINT", "BOOL", "[]BIGINT", "JSONB", "XML",
+// "TIMESTAMP".
+func makeGoLangTypeName(ti typeInfo) string {
+	switch ti.TypeId {
+	case typeInt1:
+		return "TINYINT"
+	case typeInt2:
+		return "SMALLINT"
+	case typeInt4:
+		return "INT"
+	case typeInt8:
+		return "BIGINT"
+	case typeFlt4:
+		return "REAL"
+	case typeIntN:
+		switch ti.Size {
+		case 1:
+			return "TINYINT"
+		case 2:
+			return "SMALLINT"
+		case 4:
+			return "INT"
+		case 8:
+			return "BIGINT"
+		default:
+			panic("invalid size of INTNTYPE")
+		}
+	case typeFlt8:
+		return "FLOAT"
+	case typeFltN:
+		switch ti.Size {
+		case 4:
+			return "REAL"
+		case 8:
+			return "FLOAT"
+		default:
+			panic("invalid size of FLNNTYPE")
+		}
+	case typeBigVarBin:
+		return "VARBINARY"
+	case typeVarChar:
+		return "VARCHAR"
+	case typeNVarChar:
+		return "NVARCHAR"
+	case typeBit, typeBitN:
+		return "BIT"
+	case typeDecimalN, typeNumericN:
+		return "DECIMAL"
+	case typeMoney, typeMoney4, typeMoneyN:
+		switch ti.Size {
+		case 4:
+			return "SMALLMONEY"
+		case 8:
+			return "MONEY"
+		default:
+			panic("invalid size of MONEYN")
+		}
+	case typeDateTim4:
+		return "SMALLDATETIME"
+	case typeDateTime:
+		return "DATETIME"
+	case typeDateTimeN:
+		switch ti.Size {
+		case 4:
+			return "SMALLDATETIME"
+		case 8:
+			return "DATETIME"
+		default:
+			panic("invalid size of DATETIMEN")
+		}
+	case typeDateTime2N:
+		return "DATETIME2"
+	case typeDateN:
+		return "DATE"
+	case typeTimeN:
+		return "TIME"
+	case typeDateTimeOffsetN:
+		return "DATETIMEOFFSET"
+	case typeBigVarChar:
+		return "VARCHAR"
+	case typeBigChar:
+		return "CHAR"
+	case typeNChar:
+		return "NCHAR"
+	case typeGuid:
+		return "UNIQUEIDENTIFIER"
+	case typeXml:
+		return "XML"
+	case typeText:
+		return "TEXT"
+	case typeNText:
+		return "NTEXT"
+	case typeImage:
+		return "IMAGE"
+	case typeVariant:
+		return "SQL_VARIANT"
+	case typeBigBinary:
+		return "BINARY"
+	default:
+		panic(fmt.Sprintf("not implemented makeDecl for type %d", ti.TypeId))
+	}
+}
+
+// makes go/sql type length as described below
+// It should return the length
+// of the column type if the column is a variable length type. If the column is
+// not a variable length type ok should return false.
+// If length is not limited other than system limits, it should return math.MaxInt64.
+// The following are examples of returned values for various types:
+//   TEXT          (math.MaxInt64, true)
+//   varchar(10)   (10, true)
+//   nvarchar(10)  (10, true)
+//   decimal       (0, false)
+//   int           (0, false)
+//   bytea(30)     (30, true)
+func makeGoLangTypeLength(ti typeInfo) (int64, bool) {
+	switch ti.TypeId {
+	case typeInt1:
+		return 0, false
+	case typeInt2:
+		return 0, false
+	case typeInt4:
+		return 0, false
+	case typeInt8:
+		return 0, false
+	case typeFlt4:
+		return 0, false
+	case typeIntN:
+		switch ti.Size {
+		case 1:
+			return 0, false
+		case 2:
+			return 0, false
+		case 4:
+			return 0, false
+		case 8:
+			return 0, false
+		default:
+			panic("invalid size of INTNTYPE")
+		}
+	case typeFlt8:
+		return 0, false
+	case typeFltN:
+		switch ti.Size {
+		case 4:
+			return 0, false
+		case 8:
+			return 0, false
+		default:
+			panic("invalid size of FLNNTYPE")
+		}
+	case typeBit, typeBitN:
+		return 0, false
+	case typeDecimalN, typeNumericN:
+		return 0, false
+	case typeMoney, typeMoney4, typeMoneyN:
+		switch ti.Size {
+		case 4:
+			return 0, false
+		case 8:
+			return 0, false
+		default:
+			panic("invalid size of MONEYN")
+		}
+	case typeDateTim4, typeDateTime:
+		return 0, false
+	case typeDateTimeN:
+		switch ti.Size {
+		case 4:
+			return 0, false
+		case 8:
+			return 0, false
+		default:
+			panic("invalid size of DATETIMEN")
+		}
+	case typeDateTime2N:
+		return 0, false
+	case typeDateN:
+		return 0, false
+	case typeTimeN:
+		return 0, false
+	case typeDateTimeOffsetN:
+		return 0, false
+	case typeBigVarBin:
+		if ti.Size == 0xffff {
+			return 2147483645, true
+		} else {
+			return int64(ti.Size), true
+		}
+	case typeVarChar:
+		return int64(ti.Size), true
+	case typeBigVarChar:
+		if ti.Size == 0xffff {
+			return 2147483645, true
+		} else {
+			return int64(ti.Size), true
+		}
+	case typeBigChar:
+		return int64(ti.Size), true
+	case typeNVarChar:
+		if ti.Size == 0xffff {
+			return 2147483645 / 2, true
+		} else {
+			return int64(ti.Size) / 2, true
+		}
+	case typeNChar:
+		return int64(ti.Size) / 2, true
+	case typeGuid:
+		return 0, false
+	case typeXml:
+		return 1073741822, true
+	case typeText:
+		return 2147483647, true
+	case typeNText:
+		return 1073741823, true
+	case typeImage:
+		return 2147483647, true
+	case typeVariant:
+		return 0, false
+	case typeBigBinary:
+		return 0, false
+	default:
+		panic(fmt.Sprintf("not implemented makeDecl for type %d", ti.TypeId))
+	}
+}
+
+// makes go/sql type precision and scale as described below
+// It should return the length
+// of the column type if the column is a variable length type. If the column is
+// not a variable length type ok should return false.
+// If length is not limited other than system limits, it should return math.MaxInt64.
+// The following are examples of returned values for various types:
+//   TEXT          (math.MaxInt64, true)
+//   varchar(10)   (10, true)
+//   nvarchar(10)  (10, true)
+//   decimal       (0, false)
+//   int           (0, false)
+//   bytea(30)     (30, true)
+func makeGoLangTypePrecisionScale(ti typeInfo) (int64, int64, bool) {
+	switch ti.TypeId {
+	case typeInt1:
+		return 0, 0, false
+	case typeInt2:
+		return 0, 0, false
+	case typeInt4:
+		return 0, 0, false
+	case typeInt8:
+		return 0, 0, false
+	case typeFlt4:
+		return 0, 0, false
+	case typeIntN:
+		switch ti.Size {
+		case 1:
+			return 0, 0, false
+		case 2:
+			return 0, 0, false
+		case 4:
+			return 0, 0, false
+		case 8:
+			return 0, 0, false
+		default:
+			panic("invalid size of INTNTYPE")
+		}
+	case typeFlt8:
+		return 0, 0, false
+	case typeFltN:
+		switch ti.Size {
+		case 4:
+			return 0, 0, false
+		case 8:
+			return 0, 0, false
+		default:
+			panic("invalid size of FLNNTYPE")
+		}
+	case typeBit, typeBitN:
+		return 0, 0, false
+	case typeDecimalN, typeNumericN:
+		return int64(ti.Prec), int64(ti.Scale), true
+	case typeMoney, typeMoney4, typeMoneyN:
+		switch ti.Size {
+		case 4:
+			return 0, 0, false
+		case 8:
+			return 0, 0, false
+		default:
+			panic("invalid size of MONEYN")
+		}
+	case typeDateTim4, typeDateTime:
+		return 0, 0, false
+	case typeDateTimeN:
+		switch ti.Size {
+		case 4:
+			return 0, 0, false
+		case 8:
+			return 0, 0, false
+		default:
+			panic("invalid size of DATETIMEN")
+		}
+	case typeDateTime2N:
+		return 0, 0, false
+	case typeDateN:
+		return 0, 0, false
+	case typeTimeN:
+		return 0, 0, false
+	case typeDateTimeOffsetN:
+		return 0, 0, false
+	case typeBigVarBin:
+		return 0, 0, false
+	case typeVarChar:
+		return 0, 0, false
+	case typeBigVarChar:
+		return 0, 0, false
+	case typeBigChar:
+		return 0, 0, false
+	case typeNVarChar:
+		return 0, 0, false
+	case typeNChar:
+		return 0, 0, false
+	case typeGuid:
+		return 0, 0, false
+	case typeXml:
+		return 0, 0, false
+	case typeText:
+		return 0, 0, false
+	case typeNText:
+		return 0, 0, false
+	case typeImage:
+		return 0, 0, false
+	case typeVariant:
+		return 0, 0, false
+	case typeBigBinary:
+		return 0, 0, false
 	default:
 		panic(fmt.Sprintf("not implemented makeDecl for type %d", ti.TypeId))
 	}
diff --git a/vendor/github.com/denisenkom/go-mssqldb/uniqueidentifier.go b/vendor/github.com/denisenkom/go-mssqldb/uniqueidentifier.go
new file mode 100644
index 0000000000..c8ef3149b1
--- /dev/null
+++ b/vendor/github.com/denisenkom/go-mssqldb/uniqueidentifier.go
@@ -0,0 +1,74 @@
+package mssql
+
+import (
+	"database/sql/driver"
+	"encoding/hex"
+	"errors"
+	"fmt"
+)
+
+type UniqueIdentifier [16]byte
+
+func (u *UniqueIdentifier) Scan(v interface{}) error {
+	reverse := func(b []byte) {
+		for i, j := 0, len(b)-1; i < j; i, j = i+1, j-1 {
+			b[i], b[j] = b[j], b[i]
+		}
+	}
+
+	switch vt := v.(type) {
+	case []byte:
+		if len(vt) != 16 {
+			return errors.New("mssql: invalid UniqueIdentifier length")
+		}
+
+		var raw UniqueIdentifier
+
+		copy(raw[:], vt)
+
+		reverse(raw[0:4])
+		reverse(raw[4:6])
+		reverse(raw[6:8])
+		*u = raw
+
+		return nil
+	case string:
+		if len(vt) != 36 {
+			return errors.New("mssql: invalid UniqueIdentifier string length")
+		}
+
+		b := []byte(vt)
+		for i, c := range b {
+			switch c {
+			case '-':
+				b = append(b[:i], b[i+1:]...)
+			}
+		}
+
+		_, err := hex.Decode(u[:], []byte(b))
+		return err
+	default:
+		return fmt.Errorf("mssql: cannot convert %T to UniqueIdentifier", v)
+	}
+}
+
+func (u UniqueIdentifier) Value() (driver.Value, error) {
+	reverse := func(b []byte) {
+		for i, j := 0, len(b)-1; i < j; i, j = i+1, j-1 {
+			b[i], b[j] = b[j], b[i]
+		}
+	}
+
+	raw := make([]byte, len(u))
+	copy(raw, u[:])
+
+	reverse(raw[0:4])
+	reverse(raw[4:6])
+	reverse(raw[6:8])
+
+	return raw, nil
+}
+
+func (u UniqueIdentifier) String() string {
+	return fmt.Sprintf("%X-%X-%X-%X-%X", u[0:4], u[4:6], u[6:8], u[8:10], u[10:])
+}
diff --git a/vendor/modules.txt b/vendor/modules.txt
index ef253f5194..fe21d938d3 100644
--- a/vendor/modules.txt
+++ b/vendor/modules.txt
@@ -85,8 +85,9 @@ github.com/couchbase/vellum/utf8
 github.com/couchbaselabs/go-couchbase
 # github.com/davecgh/go-spew v1.1.1
 github.com/davecgh/go-spew/spew
-# github.com/denisenkom/go-mssqldb v0.0.0-20190121005146-b04fd42d9952 => github.com/denisenkom/go-mssqldb v0.0.0-20161128230840-e32ca5036449
+# github.com/denisenkom/go-mssqldb v0.0.0-20190121005146-b04fd42d9952 => github.com/denisenkom/go-mssqldb v0.0.0-20180314172330-6a30f4e59a44
 github.com/denisenkom/go-mssqldb
+github.com/denisenkom/go-mssqldb/internal/cp
 # github.com/dgrijalva/jwt-go v0.0.0-20161101193935-9ed569b5d1ac
 github.com/dgrijalva/jwt-go
 # github.com/edsrzf/mmap-go v0.0.0-20170320065105-0bce6a688712