[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Still hang on svn 4951 RedHat 7.3 SMP

From: Philip Martin <philip_at_codematters.co.uk>
Date: 2003-02-20 12:58:43 CET

David Summers <david@summersoft.fay.ar.us> writes:

> On 19 Feb 2003, Philip Martin wrote:
> > You need to run the test manually to avoid truncating the buffered output.
> >
> > $ cd subversion/tests/clients/cmdline
> > $ ./externals_tests.py 6 BASE_URL=svn://localhost
>
> OK, Finally figured out how to run it manually. I've run it 20-30 times
> manually and frequently (but not always) it hangs. Here is the last part
> of the manual run for this one. In this particular run, I'm hanging with
> 8 lt-svnserve processes running (Ack!).

Are some of these lt-svnserve processes left over from previous hangs?
When you rerun the regression tests any old repositories get deleted
and new repositories get created. I'm not sure what would happen if
you have old lt-svnserve processes accessing the old repositories,
will they interfere with the new ones? Would you be able to try again
but this time make sure you kill any the old lt-svnserve processes and
start a new svnserve daemon after each hang.

> Here are the traces of the lt-svn and
> the 8 lt-svnserve processes and the results of manually running the
> externals_tests.py:
>
> RedHat 7.3 non-SMP
>
> externals_tests.py output (run manually):
> ==========================================
> CMD: svn "up" "working_copies/externals_tests-2.init" <TIME = 0.002711>

Last time it was test 6, now it's test 2. I assume this means that it
hangs in lots of different places.

> CMD: svnadmin "create" "repositories/externals_tests-2.other" <TIME =
> 0.002638>
> CMD: svnadmin dump "repositories/externals_tests-2" | svnadmin load
> "repositories/externals_tests-2.other" <TIME = 0.009731>
> CMD: svn "pset" "-F" "working_copies/externals_tests-2.init/tmp0mU5ZY"
> "svn:externals" "working_copies/externals_tests-2.init/A/B" <TIME =
> 0.002714>
> CMD: svn "pset" "-F" "working_copies/externals_tests-2.init/tmp0mU5ZY"
> "svn:externals" "working_copies/externals_tests-2.init/A/D" <TIME =
> 0.002633>
> CMD: svn "ci" "-m" "log msg" "working_copies/externals_tests-2.init" <TIME
> = 0.002701>
> CMD: svn "status" "-v" "-u" "-q" "working_copies/externals_tests-2.init"
> <TIME = 0.002605>
> CMD: svn "checkout" "--username" "jrandom" "--password" "rayjandom"
> "svn://localhost/repositories/externals_tests-2"
> "working_copies/externals_tests-2" <TIME = 0.002673>
> CMD: svn "checkout" "--username" "jrandom" "--password" "rayjandom"
> "svn://localhost/repositories/externals_tests-2"
> "working_copies/externals_tests-2.other" <TIME = 0.006195>
> CMD: svn "pset" "-F" "local_tmp/tmpXoZLqH" "svn:externals"
> "working_copies/externals_tests-2/A/D" <TIME = 0.002666>
> CMD: svn "ci" "-m" "log msg" "--quiet"
> "working_copies/externals_tests-2/A/D" <TIME = 0.002444>
> CMD: svn "up" "working_copies/externals_tests-2.other" <TIME = 0.002650>
>
> lt-svn
> ======
> #14 0x40024d15 in svn_client__update_internal (
> path=0x8095900 "working_copies/externals_tests-2.other/A/B/exdir_G",
> revision=0x8095818, recurse=1, timestamp_sleep=0xbfffeb48,
> ctx=0xbfffed70,
> pool=0x805deb0) at subversion/libsvn_client/update.c:138

This is the first svn:externals directory to get updated after updating
the working_copies/externals_tests-2.other directory. So this process
has just completed an update and is now failing on a second update. I
don't know what to make of this.

-- 
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Feb 20 12:59:33 2003

This is an archived mail posted to the Subversion Dev mailing list.