[Xitami] xitami aborts
Ewen McNeill
xitami@lists.xitami.org
Sat, 18 Aug 2001 13:09:33 +1200
In message <EX-20010816-171855-0001@exchange.fiserv-missive1.fiserv.net>, "Bowde
n, Russell" writes:
>This is my first posting here. I've been running Xitami with LRWP for about 2
>years on a private network doing transaction processing on Unixware 7.1,
>currently averaging about 4000 transactions per hour. My problem is that
>the Xitami process just disappears now and then.
:-(
These sorts of problems are generally caused by a bug in the software
(ie, Xitami), where it doesn't handle something properly and either
crashes on the spot, or messes something else up causing a crash later.
I see you are running Xitami 2.5b4 -- this is a beta test version. It's
still in beta testing because of issues like this (and because no one
has had time to go through and make it properly robust). While there is
a 2.5b5 out, the changes between 2.5b4 and 2.5b5 are Windows-related
(for the device name denial-of-service security issue), so are unlikely
to make much difference for you.
>Sometimes (but not always) Xitami leaves behind a memtrace.lst.
>Here's the front of one of them:
memtrace.lst is Xitami's list of memory that it knows that it's
allocated, but doesn't have any record of cleaning up properly.
Generally this means that it's exited improperly (without going through
the full close down cycle), but you know that already :-)
If your application will work with Xitam 2.4d9 (the current stable
release) you might want to consider using that. It's known to be more
stable than the beta test version, and possibly it might avoid your
problem too.
You might also want to check the console logs (Xitami, and where ever
you directed stdout/stderr of the Xitami software) to see if there are
any further details on why it chose to exit. Typically if it encounters
an internal error it does indicate waht it was doing at the time.
Ewen
--
Ewen McNeill, Technical Consultant, iMatix Corporation www.imatix.com