Samer, is your non-OpenACS "another back-end of Oracle Applications"
accessing your OpenACS Oracle schema in any way? If it is
not then AFAIK there is no possible way that it's presence in
your Oracle instance could be causing transaction locks. It might
make your Oracle instance slower, yes, and might consume other
Oracle-internal resources ("latches", etc.) which might conceivable
cause the symptoms you observe, but it certainly can't cause table or
row locking in your OpenACS schema unless it actually reads from that
schema in some way. At any rate that's not the place to start looking
for the problem.
You say during the hang you observed "some locks on some tables".
Well, what tables? And what code was taking those locks?
As others have suggested above, so far there is also little reason to
believe that database locks have anything to do with your problem.
They might, they might not. Verify that your AOLserver thread
settings are correct. Use that AOLserver nstelemetry page to check
for any suspicious AOLserver locks during the hang.
That 9 seconds taken by that scheduled proc seems suspicious. What
proc is that, and what is it doing? Instrument all your scheduled
procs if necessary in order to find out. Even it turns out to be
un-related to your hanging problem, you probably should fix it
anyway. Either speed up the scheduled proc, or give it its own thread
to run in.
