Hi. I don't have a good answer, but I'd love one.
We currently have over 110000 tickets, making free text searches extremly slow. We have done two things to try and improve the performance. The first was changing all the tables from MyISAM to InnoDb (both MySQL). Second was moving the database for OTRS from the server it was running to our main DB server. None of these measures helped much. In the version of OTRS we are running (1.2.2) the free text searching is not optimal. There is no indexing on any of the fields used, and there can't be as they're text fields and not strings. Searching with wildcards makes everything slow anyway, as indexing would not help. We have just updated the "CustomerID" field for all tickets missing this with the first used email address (inbound, outbound, note). This should let our users search for all tickets using the "CustomerID" field rather then using "From" or "To" fields in the email. What would make this ultimately better though, would be to get our internal Custimer ID numbers into OTRS, but I can't see any seamless way of doing this. I'm sure I can "hack" this in some way, but not easily. :( Best regards Are Hoel Systems Administrator Telio AS - www.telio.no -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of moe Sent: 7. mai 2005 13:33 To: otrs@otrs.org Subject: [otrs] Manually purge old/removed tickets, performance woes hi, we're running otrs 1.3.2p01 on a rather weak "AMD Sempron 2200", 512MB, IDE softraid using the postgres backend. i'm seeing serious slowdowns (requests can take up to several minutes when multiple users are working on it) with these figures: - up to 5 concurrent users - tbl article 80982 items - tbl ticket 42625 items the following performance relevant changes have been made: - Config.PM has $Self->{TicketIndexModule} = 'Kernel::System::Ticket::IndexAccelerator::StaticDB'; - i've changed the 'id' column in the article table to bigint (to make postgres use the index) - some postgres tweaking, regular vacuum analyze / vacuum full the effect was little so i'm now seeking to reduce the article count by purging old tickets from the database. thus i've created an agent job to delete tickets that are either in state 'removed' or older than 8 weeks. unfornationally that job runs _really_ slow (as in ~2 tickets per second). it seems GenericAgent is doing it the hard way (find single ticket, delete, repeat...). i really want to help out by kicking off a series of "delete from .." on the database but i don't know how. can someone please enlighten me: 1. how do i find all tickets that are 'removed' or older than 42 days? (sql) 2. from which of the 91 tables do these tickets (and refs) need to be deleted in order not to break something? thanks for any advice! best regards moe _______________________________________________ OTRS mailing list: otrs - Webpage: http://otrs.org/ Archive: http://lists.otrs.org/pipermail/otrs To unsubscribe: http://lists.otrs.org/cgi-bin/listinfo/otrs Support oder Consulting f|r Ihr OTRS System? => http://www.otrs.de/ _______________________________________________ OTRS mailing list: otrs - Webpage: http://otrs.org/ Archive: http://lists.otrs.org/pipermail/otrs To unsubscribe: http://lists.otrs.org/cgi-bin/listinfo/otrs Support oder Consulting für Ihr OTRS System? => http://www.otrs.de/
