DistributionSourcePackageRelease page still oopsing with NotOneError/LocationError

Bug #580181 reported by Ursula Junque on 2010-05-13
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Launchpad itself
High
Michael Nelson

Bug Description

As seen on OOPS-1592B957 and OOPS-1593EC1023:
  NotOneError: one() used with more than one result available

Also, OOPS-1600E1649:
  LocationError: (None, 'filesize')

We're having high counts of the latter on lpnet everyday, between 600 and 700 oopses.

According to bigjools this is related to bug 536641, Fix Released. Can be reproduced in edge and lpnet.

Related branches

Michael Nelson (michael.nelson) wrote :

It certainly looks like the same issue with a new version of java in Lucid:

https://edge.launchpad.net/ubuntu/+source/sun-java6/6.20dlj-1ubuntu3/+build/1704548

But last time this happened because the package was uploaded incorrectly to the partner component of the primary archive, but that is not the case this time:

https://pastebin.canonical.com/32299/

Nonetheless, we do seem to have a corrupt cache again, although I'm not sure in this case where the duplicate record has come from. As outlined on bug 536641, we should: "2. Ensure DistroSeries.updatePackageCache() doesn't create a new cache item for a distro archive if one already exists in another distro archive," to ensure this situation can't happen again.

Changed in soyuz:
status: New → Triaged
importance: Undecided → High
Changed in soyuz:
assignee: nobody → Michael Nelson (michael.nelson)
Ursula Junque (ursinha) on 2010-05-26
summary: - DistributionSourcePackageRelease page still oopsing with NotOneError
+ DistributionSourcePackageRelease page still oopsing with
+ NotOneError/LocationError
description: updated
Changed in soyuz:
status: Triaged → In Progress
Michael Nelson (michael.nelson) wrote :

Sorry - I didn't get to even start this before switching back to some urgent work (failure on dogfood with released code). I'll switch back and start this once the df issue is sorted.

Changed in soyuz:
status: In Progress → Triaged
Changed in soyuz:
status: Triaged → In Progress
Changed in soyuz:
milestone: none → 10.05
Changed in soyuz:
status: In Progress → Fix Committed
tags: added: qa-needstesting
Michael Nelson (michael.nelson) wrote :

Verified on staging (which is now running 9418), that both:
https://staging.launchpad.net/ubuntu/+source/sun-java6/6.20dlj-1ubuntu3/+index and
https://staging.launchpad.net/ubuntu/lucid/+source/sun-java6/6.18-2/+index

are displaying fine, while the edge counterparts oops as expected.

tags: added: qa-ok
removed: qa-needstesting
Curtis Hovey (sinzui) on 2010-06-02
Changed in soyuz:
status: Fix Committed → Fix Released
Ursula Junque (ursinha) wrote :

I still see this happening on lpnet, about 25 a day, mostly by robots. Eg.: OOPS-1617F96, OOPS-1617O124.

Michael Nelson (michael.nelson) wrote :

Thanks Ursula,

So the original oops above for the "LocationError: (None, 'filesize')" reported was OOPS-1600E1649 and was actually occurring during the distro-series source package release (corresponded to lib/lp/soyuz/templates/sourcepackagerelease-files.pt) when displaying files that have been deleted (no content). This is now tested and fixed, but is slightly different from the new oopses, which correspond to distribution source package release (in a separate template lib/lp/soyuz/templates/distributionsourcepackagerelease-index.pt) (I'd not realised there was a second point of failure).

The same fix that is tested and implemented at:
https://code.edge.launchpad.net/~michael.nelson/launchpad/580181-only-one-ds-cache/+merge/26329

needs to be applied to the distributionsourcepackagerelease template.

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

Other bug subscribers