[Evergreen-admin] Staff Client MARC Batch Import Failure

[Jeff Green]

Yes, I did increase the log level, recreate the issue, and then reviewed
the logs but I didn't see anything besides the errors copied to the
original message. Grr..

On Saturday, November 30, 2013, Sharp, Chris wrote:

> You might consider upping your loglevel for opensrf in
> /openils/conf/opensrf_core.xml.  That might provide more detail on what's
> going on.
>
> ----- Original Message -----
> > From: "Jeff Green" <jeff.green.ca at gmail.com <javascript:;>>
> > To: "Chris Sharp" <csharp at georgialibraries.org>
> > Cc: "evergreen-admin" <evergreen-admin at list.evergreen-ils.org>, "Alexey
> Vladimirovich Lazar" <alexey.lazar at mnsu.edu>
> > Sent: Friday, November 29, 2013 11:36:17 AM
> > Subject: Re: [Evergreen-admin] Staff Client MARC Batch Import Failure
> >
> >
> > Thank you Chris.
> >
> >
> > Yes, I can see files being created in /tmp including recently,
> > verified permissions on /tmp and the files is correct, etc. (we're
> > using a single server instance).
> >
> >
> > So, the MARC file gets uploaded OK, but then when vandelay goes to do
> > its thing it throws the "unable to read MARC file" error. :-/
> >
> >
> > Other ideas, folks?
> >
> >
> >
> > On Wed, Nov 27, 2013 at 10:26 AM, Sharp, Chris <
> > csharp at georgialibraries.org > wrote:
> >
> >
> > Jeff,
> >
> > The directory on the server is defined in /openils/conf/opensrf.xml
> > in the <open-ils.vandelay> section (by default it's /tmp). As the
> > inline comments in that file indicate, if you're in a multi-server
> > setup, that directory needs to be a writeable NFS share. In our case
> > it's /openils/var/data/offline/vandelay, but that's probably not a
> > standard location. If you can track down whether the files are being
> > created on the server, that might be the clue you're looking for.
> >
> > Hope that helps,
> >
> > Chris
> >
> >
> >
> > ----- Original Message -----
> > > From: "Jeff Green" < jeff.green.ca at gmail.com >
> > > To: "Alexey Vladimirovich Lazar" < alexey.lazar at mnsu.edu >
> > > Cc: "< evergreen-admin at list.evergreen-ils.org >" <
> > > evergreen-admin at list.evergreen-ils.org >
> > > Sent: Tuesday, November 26, 2013 7:54:22 PM
> > > Subject: Re: [Evergreen-admin] Staff Client MARC Batch Import
> > > Failure
> > >
> > >
> > >
> > > Hi Aleksey. Responses to your questions below.
> > >
> > >
> > > Anyone know what the mechanism is for uploading the MARC file from
> > > the staff client into the vandelay module, and what could fail
> > > during that process? When I check the relevant Perl I see that the
> > > error message is supposed to include the filename, but the log just
> > > shows null at that position, therefore I am wondering if the data
> > > is
> > > even making it to the vandelay module.
> > >
> > >
> > >
> > > We did not make any other system changes - only restarted services
> > > as
> > > it had been a while. Performance degradation over time (3-4 months)
> > > is fairly normal for us.
> > >
> > >
> > > Yes, our service restart included apache, memcache, and ejabberd as
> > > well as the Evergreen services. We have also since restarted
> > > postgres.
> > >
> > >
> > > No new errors in the postgres log, and nothing thrown during the
> > > transaction.
> > >
> > >
> > > We'll discuss the possibility of testing on a newer release to see
> > > if
> > > we can reproduce the issue there.
> > >
> > >
> > >
> > > On Mon, Nov 25, 2013 at 9:23 AM, Lazar, Alexey Vladimirovich <
> > > alexey.lazar at mnsu.edu > wrote:
> > >
> > >
> > >
> > > On 2013-11-24, at 10:45 , Jeff Green < jeff.green.ca at gmail.com >
> > > wrote:
> > >
> > > > After restarting our Evergreen services several weeks ago (due to
> > > > some slowness), we've been unable to use the MARC Batch Import
> > > > feature in



-- 
Jeff Green
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.evergreen-ils.org/pipermail/evergreen-admin/attachments/20131130/41c7c8d7/attachment.html>