This repository demonstrates how you can call IBM MQ from applications written in the Go language.
NOTICE: Please ensure that you use a dependency management tool such as dep or Glide, and add a specific version dependency.
This repository previously contained sample programs that exported MQ statistics to some monitoring packages. These have now been moved to a GitHub repository called mq-metric-samples.
A minimum level of MQ V8 is required to build these packages. However, note that
the monitoring data published by the queue manager and exploited in the mqmetric package
is not available before MQ V9. A limited set of metrics can be monitored for MQ V8 instances by setting ibmmq.usePublications=false
.
This package is provided as-is with no guarantees of support or updates. There are also no guarantees of compatibility with any future versions of the package; the API is subject to change based on any feedback. Versioned releases are made in this repository to assist with using stable APIs.
The ibmmq directory contains a Go package, exposing an MQI-like interface.
The intention is to give an API that is more natural for Go programmers than the common procedural MQI. For example, fixed length string arrays from the C API such as MQCHAR48 are represented by the native Go string type. Conversion between these types is handled within the ibmmq package itself, removing the need for Go programmers to know about it.
Sample programs are provided to demonstrate various features of using the MQI. See the
README in the samples
directory for more information about those programs.
The mqmetric directory contains functions to help monitoring programs access MQ status and statistics. This package is not needed for general application programs.
To use code in this repository, you will need to be able to build Go applications, and
have a copy of MQ installed to build against. It uses cgo to access the MQI C
structures and definitions. It assumes that MQ has been installed in the default
location (on a Linux platform this would be /opt/mqm
) but this can be changed
with environment variables if necessary.
Windows compatibility is also included. This has been tested with Go 1.10 compiler, which now permits standard Windows paths (eg including spaces) so the CGO directives can point at the normal MQ install path.
If you are unfamiliar with Go, the following steps can help create a working environment with source code in a suitable tree. Initial setup tends to be platform-specific, but subsequent steps are independent of the platform.
-
Install the Go runtime and compiler. On Linux, the packaging may vary but a typical directory for the code is
/usr/lib/golang
. The compiler should be at least version 10. If you see an error similar to "ld: NULL not defined" when building a program then it is likely you need to upgrade your compiler. -
Create a working directory. For example,
mkdir $HOME/gowork
-
Set environment variables. Based on the previous lines,
export GOROOT=/usr/lib/golang
export GOPATH=$HOME/gowork
- On Linux, you must set environment variables to permit some compile/link flags. This is due to a security fix in the compiler.
export CGO_LDFLAGS_ALLOW="-Wl,-rpath.*"
- Install the git client
- Install the Go runtime and compiler. On Windows, the common directory is
c:\Go
- Ensure you have a gcc-based compiler, for example from the Cygwin distribution.
I recommend you use the mingw variation, to ensure compiled code can be used on systems without Cygwin installed.
The default
gcc
compiler from Cygwin does not work because it tries to build a Cygwin-enabled executable but the MQ libraries do not work in that model; themingw
versions build Windows-native programs. - Create a working directory. For example,
mkdir c:\Gowork
- Set environment variables. Based on the previous lines,
set GOROOT=c:\Go
set GOPATH=c:\Gowork
set CC=x86_64-w64-mingw32-gcc.exe
- The
CGO_LDFLAGS_ALLOW
variable is not needed on Windows - Install the git client
-
Make sure your PATH includes routes to the Go compiler (
$GOROOT/bin
), the Git client, and the C compiler. -
Change directory to the workspace you created earlier. (
cd $GOPATH
) -
Use git to get a copy of the MQ components into a new directory in the workspace.
git clone https://github.com/ibm-messaging/mq-golang.git src/github.com/ibm-messaging/mq-golang
-
If you have not installed MQ libraries into the default location, then set environment variables for the C compiler to recognise those directories. You may get messages from the compiler saying that the default MQ directories cannot be found, but those warnings can be ignored. The exact values for these environment variables will vary by platform, but follow the corresponding CFLAGS/LDFLAGS values in
mqi.go
For example,
export MQ_INSTALLATION_PATH=/my/mq/dir # This will also be set from the setmqenv command
export CGO_CFLAGS="-I$MQ_INSTALLATION_PATH/inc"
export CGO_LDFLAGS="-L$MQ_INSTALLATION_PATH/lib64 -Wl,-rpath,$MQ_INSTALLATION_PATH/lib64"
-
Compile the
ibmmq
component:go install ./src/github.com/ibm-messaging/mq-golang/ibmmq
-
If you plan to use monitoring functions, then compile the
mqmetric
component:go install ./src/github.com/ibm-messaging/mq-golang/mqmetric
-
Sample programs can be compiled in this way
go build -o bin/mqitest ./src/github.com/ibm-messaging/mq-golang/samples/mqitest/*.go
At this point, you should have a compiled copy of the program in $GOPATH/bin
. See the
samples
directory for more sample programs.
The buildSamples.sh
script in this directory can also be used to create a container which will
compile the samples and copy them to a local directory.
These GitHub-hosted projects are related to or derived from this one. This is not a complete list. Please let me know, via an issue, if you have another project that might be suitable for inclusion here.
Repository | Description |
---|---|
ibm-messaging/mq-metric-samples | Extracts metrics for use in Prometheus, Influx JSON consumers etc. |
ibm-messaging/mq-golang-jms20 | JMS-style messaging interface for Go applications |
ibm-messaging/mq-container | Building MQ into containers. Uses features from this package for configuration and monitoring |
felix-lessoer/qbeat | Extract monitoring and statstics from MQ for use in Elasticsearch |
ibm-messaging/mq-mqi-nodejs | A similar MQI interface for Node.js applications |
All regular MQI verbs are now available through the ibmmq
package.
The only unimplemented area of MQI function is the use of Distribution Lists: they were rarely used, and the Publish/Subscribe operations provide similar capability.
- There is currently a queue manager limitation which does not permit resource publications to be made about queues whose name includes '/'. Attempting to monitor such a queue will result in a warning logged by the mqmetric package.
See CHANGELOG in this directory.
Feedback on the utility of this package, thoughts about whether it should be changed or extended are welcomed.
For feedback and issues relating specifically to this package, please use the GitHub issue tracker.
Contributions to this package can be accepted under the terms of the IBM Contributor License Agreement, found in the CLA file of this repository. When submitting a pull request, you must include a statement stating you accept the terms in the CLA.
© Copyright IBM Corporation 2016, 2019