Product:+filebug timeouts

Bug #722787 reported by Ursula Junque
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Launchpad itself
Fix Released
Critical
Unassigned

Bug Description

 OOPS-1875O1575, OOPS-1875O1569
https://bugs.launchpad.net/ubuntu-tweak/+filebug

Revno: 12399
SQL time: 15359 ms
Non-sql time: 247 ms
Total time: 15606 ms
Statement Count: 29

 15038.0 1 SQL-launchpad-main-master
SELECT BugTask.assignee, BugTask.bug, BugTask.bugwatch, BugTask.date_assigned, BugTask.date_closed, BugTask.date_confirmed, BugTask.date_fix_committed, BugTask.date_fix_released, BugTask.date_incomplete, BugTask.date_inprogress, BugTask.date_left_closed, BugTask.date_left_new, BugTask.date_triaged, BugTask.datecreated, BugTask.distribution, BugTask.distroseries, BugTask.id, BugTask.importance, BugTask.milestone, BugTask.OWNER, BugTask.product, BugTask.productseries, BugTask.sourcepackagename, BugTask.status, BugTask.statusexplanation, BugTask.targetnamecache
FROM BugTask, Bug
WHERE Bug.id = BugTask.bug
  AND BugTask.product = $INT
  AND Bug.fti @@ ftq($STRING)
  AND (Bug.private = FALSE
       OR EXISTS
         (SELECT BugSubscription.bug
          FROM BugSubscription,
               TeamParticipation
          WHERE TeamParticipation.person = $INT
            AND TeamParticipation.team = BugSubscription.person
            AND BugSubscription.bug = Bug.id
          UNION SELECT BugTask.bug
          FROM BugTask,
               TeamParticipation
          WHERE TeamParticipation.person = $INT
            AND TeamParticipation.team = BugTask.assignee
            AND BugTask.bug = Bug.id))
ORDER BY -rank(Bug.fti, ftq($STRING)), BugTask.bug LIMIT $INT
OFFSET $INT

with a nuts long search string. Its not clear why we do that search during the filebug POST.

Tags: timeout
description: updated
Revision history for this message
j.c.sackett (jcsackett) wrote :

I believe this is the step wherein it's trying to detect bugs that are similar to the one you're reporting, in which case the real culprit is bug 716786

Revision history for this message
Robert Collins (lifeless) wrote :

Was due to the fat indices - fixed by the recreation this week.

Changed in launchpad:
status: Triaged → Fix Released
Revision history for this message
Stuart Bishop (stub) wrote :

If you look at the OOPS reports, the queries without parameter substitution are slamming insanely huge strings of boolean into the ftq() function. Even if the query runs fast, we will be getting garbage out because we are putting garbage in.

Revision history for this message
Robert Collins (lifeless) wrote :

So the booleans are doing the N terms-1 permutations of the text put in, it works pretty well (better than a plain search, faster than the abomination we had before). We need a per-context document term frequency filter to do better (which things like lucene just build in, but tsearch2 doesn't seem to have).

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.