Product:+packages timesout with no packages and one series
Bug #663857 reported by
Curtis Hovey
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Launchpad itself |
Fix Released
|
Critical
|
Edwin Grubbs |
Bug Description
OOPS-1754F409 https:/
(newer oops)
OOPS-1773EA141
1. This page should not be accessible to the code vhost.
2. This project has no packages (and delete series ensures that)
3. There is only one active series.
This page is timing out as it iterates over hundred of obsolete series that do not have packages,
Branch: launchpad-rev-11858
Revno: 11858
SQL time: 12866 ms
Non-sql time: 2765 ms
Total time: 15631 ms
Statement Count: 1319
@ 1300 queries, we need 10ms performance to avoid death by sql. Changing the dimension from packages to 'sets to query' should fix this quite handily.
Related branches
lp:~edwin-grubbs/launchpad/bug-663857-product-packages-timeout
- Curtis Hovey (community): Approve (code ui)
-
Diff: 308 lines (+86/-59)5 files modifiedlib/lp/registry/browser/product.py (+4/-19)
lib/lp/registry/browser/tests/packaging-views.txt (+19/-20)
lib/lp/registry/interfaces/product.py (+3/-0)
lib/lp/registry/model/product.py (+27/-7)
lib/lp/registry/templates/product-packages.pt (+33/-13)
lp:~edwin-grubbs/launchpad/bug-663857-product-packages-timeout-part2
- Tim Penhey (community): Approve (code)
- Curtis Hovey (community): Approve (code)
- Steve Kowalik (community): Approve (code*)
-
Diff: 106 lines (+32/-16)4 files modifiedlib/lp/registry/browser/product.py (+9/-13)
lib/lp/registry/interfaces/product.py (+2/-0)
lib/lp/registry/model/product.py (+16/-0)
lib/lp/registry/templates/product-packages.pt (+5/-3)
Changed in launchpad: | |
assignee: | nobody → Edwin Grubbs (edwin-grubbs) |
milestone: | none → 11.02 |
status: | Triaged → In Progress |
Changed in launchpad: | |
importance: | High → Critical |
Changed in launchpad: | |
status: | Fix Committed → Fix Released |
To post a comment you must log in.
Hmm, that oops is from +milestones. Odd.
Perhaps 1) should be a different bug?