Z39.50 - internal server error
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
Hi Radek, Which version of Koha are you running on? What are you local changes? Did you try without? Cheers, Jonathan On Mon, 14 Nov 2016 at 08:49 <radek.siman@centrum.cz> wrote:
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
Hi Jonathan, I think it's the latest version available: 16.05.04.001. The local changes are focused on authorities searching as I think, but I'm not absolutely sure (isn't my work). It's just my opinion that about the local changes, so we didn't try to run without it - in addition it's a bit difficult to turn it off on production systems. I would appreciate any information regarding plack, which might be missed during our development process. Is there any reason why "undefined subroutine" error raises because of plack? PLease remember, the restart always fixes the problem as well as prefixing the procedure name. But the question is: WHY? Cheers, Radek Dne 14.11.16 v 12:40 Jonathan Druart napsal(a):
Hi Radek,
Which version of Koha are you running on? What are you local changes? Did you try without?
Cheers, Jonathan
On Mon, 14 Nov 2016 at 08:49 <radek.siman@centrum.cz <mailto:radek.siman@centrum.cz>> wrote:
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 <mailto:Koha@lists.katipo.co.nz> https://lists.katipo.co.nz/mailman/listinfo/koha
El lun., 14 nov. 2016 a las 9:28, Radek Šiman (<radek.siman@centrum.cz>) escribió:
development process. Is there any reason why "undefined subroutine" error raises because of plack? PLease remember, the restart always fixes the problem as well as prefixing the procedure name. But the question is: WHY?
Probably related to forking ? -- Tomás Cohen Arazi Theke Solutions (https://theke.io <http://theke.io/>) ✆ +54 9351 3513384 GPG: B2F3C15F
Greetings, My guess is that the local changes probably create a circular use. Which is why the qualification with C4:... is necessary. Did the local changes include any use statements? :) GPML, Mark Tompsett
participants (5)
-
Jonathan Druart -
Mark Tompsett -
Radek Šiman -
radek.siman＠centrum.cz
-
Tomas Cohen Arazi