Previous Topic Next topic Print topic


Dumps and Traces

We recommend that you always run a production system with some diagnostics switched on. This is necessary so that you can troubleshoot unexpected problems. Diagnostics consume resources, and there is always contention between problem determination facilities and performance. You need to decide how important it is to be able to diagnose a problem the first time it occurs rather than having to try to recreate it or having to just turn on diagnostics and wait for it to reoccur. We suggest the following level of diagnostics as a minimum:

You might also need to turn on tracing for task control, storage control and application container components.

You specify settings for traces and internally triggered dumps on the Edit Server > Diagnostics page or the Control page of ESMAC. The trace flags on the ESMAC Control page that are equivalent to the trace flags on the Edit Server > Diagnostics page are as follows:

Edit Server > Diagnostics page ESMAC Control page
Application API
Task Control KCP
Storage Control SCP
Application Container RTS

Settings you make on the Edit Server > Diagnostics page override those you make on the ESMAC Control page.

Dumps created when the system or a transaction abends are internally triggered dumps; as long as you have selected them on the Control page, they will be created when abends occur. You can also obtain an externally triggered dump, that is, a dump that is taken immediately in response to a command. There are several ways of initiating an externally triggered dump:

The casdump command gives you more flexibility.

First run casdump as follows:

casdump /res-name

By default, casdump attempts to lock shared memory before creating the dump. If enterprise server fails in a way that leaves shared memory locked, the command might hang.

If the command hangs, you can run it again with the /d option:

casdump /res-name /d

Using /d means "do not lock shared memory". This could also cause the command to fail due to shared memory being changed by the running enterprise server while casdump is chasing and dumping storage chains.

Alternatively, use the /b option which dumps all of shared memory as a block at the very beginning before it starts to chase chains and format blocks.

If you are planning to send the dump file to Micro Focus for investigation, you might also be asked to run the command with the /f option; this option dumps the FAQE chain (Free Area Queue Element) and shared memory as a block.

An externally triggered dump is written to the dump X dataset, casdumpx.rec, if you specify the /d option, or the current dump dataset casdumpa.rec or casdumpb.rec if you do not specify it.

Previous Topic Next topic Print topic