On 19 December 2013 11:36, Galen Charlton <gmc@esilibrary.com> wrote:
Hi,
On Wed, Dec 18, 2013 at 8:28 AM, Paul Poulain <paul.poulain@biblibre.com> wrote:
Le 10/12/2013 23:50, Brendan Gallagher a écrit :
Finishing the Plackifying of the staff-side so that it could be stably used. The SAN-OP is running Plack on staff since 3.10, it runs fine, except for C4/BackgroundJobs, and we could not understand where the problem comes from, because we can't reproduce it outside from SAN-OP server (which is expected to be a Wheezy-with-nothing-exotic)
I think there may be a bit more to it considering the existence of things like bug 9224 and the lack of a consensus Plack configuration in the source tree. That said, I would agree that Plack-compatibility for the staff interface is both achievable and not far off for 3.16.
[1] http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=9224
Indeed, I have just found a bug that effects both the staff and public interface. After time a plack thread can lose its Zebra connection, meaning searches fail until that thread dies and respawns (but intermittently because some other threads might be ok). Working on a fix for this now Chris