Disable triggers when performing bib visibility recalculations at the end of the 3.0.2-3.0.3 upgrade script

[Kathy Lussier]

At the end of the 3.0.2-3.0.3 upgrade script, we perform a bib visibility recalculation that is similar to the one included in the 2.12.6 to 3.0.0 upgrade script. In the 2.12.6-3.0.0 script, we had disabled many triggers before performing the recalculation because they were unnecessary and took a very long time to perform.

Back in February, I asked a question in IRC about whether we should make similar changes for the new recalculation, and the answer was 'yes.' - http://irc.evergreen-ils.org/evergreen/2018-02-02#i_344301

Unfortunately, I never filed the bug. I'm doing so now.

See related list e-mail at https://georgialibraries.markmail.org/thread/gzx36cndjabhf4o4

[Kathy Lussier (klussier)]

Working branch at http://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/kmlussier/lplp1758426-speed-up-bre-visibility-computation-part-deux

I basically copied what was done at the end of the 2.12.6-3.0.0 upgrade script.

[Josh Stompro (u-launchpad-stompro-org)]

Seemed to work fine for me, went from 3 + hours(I gave up) to 20 seconds updating 79K bib records on a slow test system with only 8GB of ram.
Josh

[Ben Shum (bshum)]

Fix pushed to master and backported to both rel_3_1 and rel_3_0 for more efficient upgrades.