Last modified: 2014-07-18 12:37:09 UTC

Wikimedia Bugzilla is closed!

Wikimedia migrated from Bugzilla to Phabricator. Bug reports are handled in Wikimedia Phabricator.
This static website is read-only and for historical purposes. It is not possible to log in and except for displaying bug reports and their history, links might be broken. See T43612, the corresponding Phabricator task for complete and up-to-date bug report information.
Bug 41612 - Page moves of files on commons slow
Page moves of files on commons slow
Status: RESOLVED WORKSFORME
Product: Wikimedia
Classification: Unclassified
Media storage (Other open bugs)
wmf-deployment
All All
: Low normal (vote)
: ---
Assigned To: Nobody - You can work on this!
aklapper-moreinfo
: performance
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2012-11-01 00:36 UTC by Smallman
Modified: 2014-07-18 12:37 UTC (History)
9 users (show)

See Also:
Web browser: ---
Mobile Platform: ---
Assignee Huggle Beta Tester: ---


Attachments

Description Smallman 2012-11-01 00:36:09 UTC
For the commons, it seems that page moves for files (done from the api) are very slow...is this normal?
Comment 1 Andre Klapper 2012-11-02 11:42:15 UTC
Smallman: Which exact API command do you use? 
What does "very slow" mean, as this is rather subjective?
Comment 2 Smallman 2012-11-03 00:52:38 UTC
By slow, I meant 10+ seconds. The files moved were pdfs 3+ MB in size.

The files were moved by smallbot on the commons:
https://commons.wikimedia.org/wiki/Special:Contributions/Smallbot

in the category:
https://commons.wikimedia.org/wiki/Category:Gerald_R._Ford_Presidential_Library_and_Museum_series:_Budget_Review_Decision_Papers

I used:
HTTP POST w/ "application/x-www-form-urlencoded"
https://commons.wikimedia.org/w/api.php?action=move

from=<from>&to=<to>&token=<token>&reason=<reason>&noredirect=&format=xml

Is data (blobs) actually being copied over in a move, or is the identifier the only thing being changed?
Comment 3 Bawolff (Brian Wolff) 2012-11-05 22:06:44 UTC
I'm going to cc Aaron Schulz because he's involved with swift (I think), and this sounds likely to be a swift issue.
Comment 4 Aaron Schulz 2012-11-06 02:57:21 UTC
Note that we are still writing to swift+NFS synchronously, so it's slower than just writing to one alone. Swift also has high PUT latency, and to less extent DELETE latency.
Comment 5 Aaron Schulz 2014-05-14 22:16:49 UTC
How common is 10+ seconds now? Multiwrite has since been turned off.
Comment 6 Andre Klapper 2014-06-16 12:40:30 UTC
Smallman: Can you answer comment 5, please?
Comment 7 Andre Klapper 2014-07-18 12:37:09 UTC
Unfortunately closing this report as no further information has been provided.

Smallman: Please feel free to reopen this report if you can provide the information asked for in comment 5 now that Multiwrite is turned off, and if this still happens. Thanks!

Note You need to log in before you can comment on or make changes to this bug.


Navigation
Links