[Design of JBoss internal QA (Test Suite)] - transaction (crash recovery) tests.
by jhalliday
Hi all
I know this is a long shot, but is there any documentation for the test suite framework? I'm interested in the mechanisms used to e.g. configure, start and stop the app server.
To provide some background, I'm looking at how to approach the problem of testing transactions crash recovery in the app server. This has not been an issue until fairly recently, as the old tm did not support crash recovery anyhow. Hence there is almost certainly no support for it currently.
JBossTS has an extensive crash recovery test suite, built on the DTF framework. The framework provides for configuring, starting and stopping processes, just like the AS test suite. But it also supports configuration for external resources, mainly the various XA aware databases we test against. It also supports crashing (rather than shutting down) processes at various points. Thus we have fairly good automated test coverage for JBossTS standalone edition.
Crash recovery for JBossTS inside the app server requires different configuration and pretty clearly can't be run without the app server. The choices seem to be: drive these tests through the DTF, perhaps by writing additional plugins to support the app server configure/start/stop/crash, or drive the tests through the app server test suite.
There is a certain amount of less technical discussion to be had regarding where, conceptually speaking, these tests should reside. Are they part of the AS or part of the transaction service? Should they still be expected to pass if JBossTS is swapped out for another transactions engine, as is the case with existing tm tests in the AS test suite? Can they link against JBossTS internal classes, or only the integration SPI?
Mainly though I'm initially interested in determining how much help the existing AS test suite framework may be in providing a starting point for the functionality I need to e.g. wire up databases to the AS, crash and restart the AS, etc.
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4123552#4123552
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4123552
16 years, 11 months
[Design the new POJO MicroContainer] - Re: Context ClassLoader - POJO
by adrian@jboss.org
"alesj" wrote :
| btw: the installs have different CL usage as callbacks
| Installs have it per install instance, where callbacks are grouped in a single set.
| Why the difference?
I thought I explained it above? Let me reword it :-)
* The classloader usage is the same from the user's point of view.
e.g. The addDeployer() invocation on the MainDeployer will have the context classloader
of the Deployer being added regardless of which mechanism is used.
It's different from our point of view because install and incallback are implemented
in different ways.
* I didn't understand how the callbacks hang together so I just wrapped all the
processing inside the classloader change. If it causes problems then
we'll (most likely you :-) will have to workout how to do it more fine-grained.
I also assumed this would take care of the callbacks like the class attribute
that I didn't know existed, let alone understand.
View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4123546#4123546
Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4123546
16 years, 11 months