https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=18428 Barton Chittenden <barton@bywatersolutions.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|normal |major --- Comment #1 from Barton Chittenden <barton@bywatersolutions.com> --- The underlying problem is that innodb doesn't keep auto_increment across SQL server restarts -- it simply sets auto_increment to max(reserves.reserve_id)+1 on startup. Because the action of moving rows from reserves to old_reserves makes old_reserves.reserve_id higher than reserves.reserve_id, rebooting the server when it's in that state will create duplicate reserve_ids between reserves and old_reserves. since we treat deleted* tables the same way, we should also fix select distinct table_name from information_schema.columns where table_name like 'deleted%'; +--------------------+ | table_name | +--------------------+ | deletedbiblio | | deletedbiblioitems | | deletedborrowers | | deleteditems | +--------------------+ -- You are receiving this mail because: You are watching all bug changes.

https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=18428 Jonathan Druart <jonathan.druart@bugs.koha-community.org> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jonathan.druart@bugs.koha-c | |ommunity.org --- Comment #2 from Jonathan Druart <jonathan.druart@bugs.koha-community.org> --- To me the only clean and safe way to do that is to merge the tables. I plan to work on that. -- You are receiving this mail because: You are watching all bug changes.

https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=18428 --- Comment #3 from Barton Chittenden <barton@bywatersolutions.com> --- (In reply to Jonathan Druart from comment #2)

To me the only clean and safe way to do that is to merge the tables. I plan to work on that.

I don't disagree that merging the tables is probably the right thing to do, but I think that we *can* safely fix the problem by adjusting the auto_increment at startup, as mentioned in IRC: 21:22 barton rangi: I think the right place to fix this is in the mysql init file. 21:22 ... we fix the problem on server startup, then we don't have to worry about it. 21:24 rangi thats not so hard for debian 21:24 there is a /etc/mysql/conf.d/ 21:24 barton by 'fix the problem' I mean set the auto_increment for issues and reserves. 21:25 rangi and my.cnf does includedir /etc/mysql/conf.d/ 21:25 so you can put a koha.cnf in there 21:25 and it will be used 21:25 and not overwritten by mysql upgrades 21:25 barton excellent. We should probably do the same thing for the deleted* tables. ... all told, I think that the tables that need to be merged or twiddled at startup is issues/old_issues, reserves/old_reserves, biblios/oldbiblios, biblioitems/oldbiblioitems, items/olditems, borrowers/deletedborrowers -- that's a lot fix at once. If we figure out how to fix one table in koha.cnf, we can do all six, then work on the plumbing. Just my two cents. -- You are receiving this mail because: You are watching all bug changes.

https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=18428 Jonathan Druart <jonathan.druart@bugs.koha-community.org> changed: What |Removed |Added ---------------------------------------------------------------------------- See Also| |https://bugs.koha-community | |.org/bugzilla3/show_bug.cgi | |?id=18931 -- You are receiving this mail because: You are watching all bug changes.

https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=18428 jplevyak@gmail.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jplevyak@gmail.com -- You are receiving this mail because: You are watching all bug changes.