Hi,
sometimes we are facing very annoying problem concerning Z39.50 searching, bacause all attempt to
search records fail with a "white page of death" saying only "Internal server
error". It is somehow connected to plack, but we cannot find a good solution.
Symptoms:
# tail plack-error.log -n 5
Undefined subroutine &C4::Breeding::GetZ3950BatchId called at
/usr/share/koha/lib/C4/Breeding.pm line 279.
Undefined subroutine &C4::Breeding::GetZ3950BatchId called at
/usr/share/koha/lib/C4/Breeding.pm line 279.
Undefined subroutine &C4::Breeding::GetZ3950BatchId called at
/usr/share/koha/lib/C4/Breeding.pm line 279.
Undefined subroutine &C4::Breeding::GetZ3950BatchId called at
/usr/share/koha/lib/C4/Breeding.pm line 279.
Undefined subroutine &C4::Breeding::GetZ3950BatchId called at
/usr/share/koha/lib/C4/Breeding.pm line 279.
Solution A:
# koha-plack --restart
When we restart the plack service, then the searching starts to work again for
an unpredictable period of time (often several days, sometimes week or two).
Solution B:
We add a "C4::ImportBatch::" prefix to C4/Breeding.pm at these two lines:
my $batch_id = GetZ3950BatchId($servhref->{servername});
my $breedingid = AddBiblioToBatch($batch_id, $seq, $marcrecord, 'UTF-8', 0, 0);
This solution works well except the fact, that prefixing is not probably the
best idea because the procedures SHOULD be already imported. In fact, they
aren't. Sometimes.
We think it could be caused by our local changes of code. We are not happy with it, but
is it possible that the local changes of code don't work well with plack module? What
should be kept in mind to avoid (probably?) incorrect imports with "use"
command?
Thanks a lot!
Radek
_______________________________________________
Koha mailing list http://koha-community.org
Koha@lists.katipo.co.nz
https://lists.katipo.co.nz/mailman/listinfo/koha
