In many cases, there is not a shutdown hook registered when there could be.
Reggie, Mahalo and Norm in particular, having "subscription" based use, should
notify their users immediately apon catchable VM shutdown. The mechanism
provided by
Runtime.getRuntime().addShutdownHook(Runnable)
should be used everywhere we can to "speed up" the notification of process
failures.
Gregg Wonderly
On 11/26/2010 8:21 AM, Patricia Shanahan wrote:
Sim IJskes - QCG wrote:
On 26-11-10 14:42, Patricia Shanahan wrote:
2. Ensuring that the River features that are used during the current
teardown are tested in some other way.
But only if the teardown is already part of the testresult. Is it?
Whether or not a teardown failure gets reflected in the test result, we
would learn about it soon enough under current conditions, due e.g. to
non-running of subsequent tests. I suspect we do have the formal test
coverage, but I'll take a look at what teardown uses now, and check.
What worries me a little is the state the operating system keeps on terminated
processes. We know of the SO_REUSEADDR issue, but you seem to experience the
non-release of resources on process termination. formally a OS bug, but if so
is something we need to workaround. How definate are you on your reports where
you have problems rerunning a test caused by the OS not releasing resources on
VM exit?
I see two issues here:
1. Terminating every VM.
2. Deleting any temporary files.
Empirically, a Ctrl-C termination often leaves the system in a state in which
subsequent tests fail.
Patricia
