cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

ActiveGate MQ Monitoring the same MQ QMgr across several servers

Larry_S
Helper

Hello,

I'm new to Dynatrace monitoring and using the ActiveGate to monitor MQSeries is new at my company.

My "basic" (at least I hope so) question is...

Is it possible to monitor the same-named MQ Queue Manager that exists on multiple servers that is configured in an Active/Standby configuration?

The way we have configured the Dynatrace ActiveGate we have defined 3 endpoints with hostname-specific names [QMgrName-HostName] but the Dashboard only shows one instance/endpoint.

I am well aware that with (for example that we have a single QMgr configured in Active/Passive across 3 nodes) that the 2 "Standby" node's QMgrs will not return data from Dynatrace polling, but we still wan to be able to see on which node the Active queue manager is...um...Active.

Does anyone else have this similar configuration?  We do not (and in Kubernetes) can not deploy agents locally on the servers and have to use ActiveGate.

Any help or acknowledgement will be greatly appreciated.

4 REPLIES 4

Karolina_Linda
Community Team
Community Team

Hello Larry,

 

Welcome to the Dynatrace Community! 👋 Thank you for all the insights.

 

I hope there's already someone that have experienced similar situation and can jump to help here.

 

Have a great day and a great time in the Dynatrace Community!

Karolina

Keep calm and build Community!

diego_morales
Dynatrace Advisor
Dynatrace Advisor

Hi Larry,

 

To answer your question whether it is possible to monitor the same-named Queue Manager that exists on multiple servers, the answer is Yes.

On the endpoint configuration, it prompts to enter a comma-separated list of Hosts and Ports, so for example: 192.168.10.11:1415, 192.168.10.12:1415. This is a connection name list which gets passed to the connection.

Keep in mind that it needs to timeout rather quickly before it goes to the next connection in the name list, so depending which extension you are running, you may have to change the TCP Stanza in the mqclient.ini to add a connection timeout value. The legacy custom.remote.python.ibmmqpcf would require this TCP Stanza change, the newer custom.remote.python.ibmmq_java one does not.

 

While the extension connects round-robin to the Active one in a multi-instance queue manager, it will not tell you which instance it connected to.

 

Hope this helps.

 

Thanks,

 

Diego

Hi Diego-

Yes, we have created a single endpoint, configured with our (3 in our configuration) MQ node hostnames and it seems to resolve to a "blue" (ok?) condition in the Dynatrace dashboard.  We were hoping for something more in the way of which, of the three host names, that Dynatrace Active Gate is able to connect to, and gather data from.  Our current mqclient.ini was not modified therefore the timeout value is of the default setting.    How can I determine, from our Dynatrace installation if we are legacy or "newer".  I can only assume we are using "newer", at this point and therefore, from your comment, will not need this mqclient.ini consideration.

Hi Larry,

 

By the name of the extension: custom.remote.python.ibmmqpcf or custom.remote.python.ibmmq_java. The former is the legacy, the latter is the newer one. Also by the version number, if it starts with 1.x, then it is the legacy one. If it starts with 2.x then it is the newer one.

 

Unfortunately the extension does not know or show you which host it connected to. It concentrates on the queue manager and is agnostic to the host it runs on.

 

Thanks,

Diego

Featured Posts