Monitoring alerts when Broker or RPC servers go down

Hello,

I am hoping someone can share best practices on how you do alert monitoring to notify someone that a Broker or an RPC server that should be registered has gone down.

Not that it happens often — SAG products are reliable — but it can and does happen for various reasons and it has a serious business impact when it does. I’d like to get that alert before a business user has to tell me through the help desk. Started tasks don’t get picked up by CA-7 either (which the Broker and Natural RPC Servers are), and Windows services on remote virtual machines are limited as well.

I am hoping there are some good / best practices someone is willing to share.

If I could have something that ran every 5 minutes and pinged to check to see if the Broker is up and if all the services expected are registered to it, and if not, then I can send an email and terminate with a code that command operations WILL call out for, but I don’t know a good way to do that. Or maybe there is a better idea?

Thanks in advance!

-Brian Johnson

Ha! I might answer myself here… I see USR2073N that I will play with here to see if that works.

You can use ETBINFO tool - execute every x minutes -

  1. to see that Broker is up and healthy - upon normal response
  2. To obtain number of RPC servers for one service

You can then start additional RPC server automatically as result of this check
This scripting is done at Migdal Insurance on Unix, comparable actions of mainframe EXX Broker.

Shmuel

You can also wrap calls to the ETBCIS services for very fine-grained control over what services are being monitored. With the RPCCIS services available, this is easier to do that it used to be. (http://techcommunity.softwareag.com/ecosystem/documentation/webmethods/wmsuites/wmsuite9-7/EntireX/9-7_EntireX/rpc/cis-idl.htm#cis-idl).

Check out the scripting that is also available for monitoring: http://techcommunity.softwareag.com/ecosystem/documentation/webmethods/wmsuites/wmsuite9-7/EntireX/9-7_EntireX/scripting/monitorEnv.htm

Hi Douglas,

I really appreciate this information. This is the most in-depth monitoring process suggested and says it will handle ensuring RPC servers are running and not just registered with the Broker so it is ideal. We can plug this into our automated monitoring alert systems, too. This is the best-of-class solution!

Regards,

Brian