Hi
I have a Subversion test scenario that fails more often than not when
the server (Apache, apr, mod_dav_svn) is built with APR pool debugging
enabled. Without pool debugging it has not yet failed.
Are there any known problems with APR pool debugging? I realise that
pool debugging will change memory usage, CPU usage, timing, cache
footprint, etc., all of which are enough to explain the difference,
but I thought I would check if there are known problems before
spending too much time investigating the Subversion bug.
The error is a Subversion Berkeley DB error when running a checkout
and commit simultaneously, either:
[Sat Oct 04 14:31:05 2003] [error] [client 127.0.0.1] Provider encountered an error while
streaming a REPORT response. [500, #0]
[Sat Oct 04 14:31:05 2003] [error] [client 127.0.0.1] A failure occurred while driving the
update report editor [500, #160014]
[Sat Oct 04 14:31:05 2003] [error] [client 127.0.0.1] reference to non-existent node '(null).0.1'
in filesystem '/home/pm/sw/subversion/obj/repostress/db' [500, #160014]
or:
[Sat Oct 04 14:31:37 2003] [error] [client 127.0.0.1] Could not MERGE resource "/obj/repostress/!svn/act/1add337e-dcc8-0310-9bef-f477c27520b2"
into "/obj/repostress/trunk". [500, #0]
[Sat Oct 04 14:31:37 2003] [error] [client 127.0.0.1] could not process the merge delta.
[500, #160014]
[Sat Oct 04 14:31:37 2003] [error] [client 127.0.0.1] (2)No such file or directory: reference
to non-existent node '(null).0.4' in filesystem '/home/pm/sw/subversion/obj/repostress/db'
[500, #160014]
I am using the prefork MPM on Linux, and Subversion HEAD (r7287). The
error occurs on both single processor and dual processor machines, and
with both Apache 2.0.47 and STRIKER_2_0_48_PRE4.
--
Philip Martin
|