Page MenuHomePhabricator

Fix multistream download link in recent dumps index.html pages
Closed, ResolvedPublic

Description

Reported to the xmldatadumps-l emailing list: https://lists.wikimedia.org/pipermail/xmldatadumps-l/2020-April/001527.html

This looks like it's due to the refactor of file listing methods done recently.

Event Timeline

ArielGlenn triaged this task as Medium priority.Apr 6 2020, 7:46 AM
ArielGlenn created this task.

Change 586308 had a related patch set uploaded (by ArielGlenn; owner: ArielGlenn):
[operations/dumps@master] fix up link for multistream index file download

https://gerrit.wikimedia.org/r/586308

Change 586308 merged by ArielGlenn:
[operations/dumps@master] fix up link for multistream index file download

https://gerrit.wikimedia.org/r/586308

The above change has been deployed to all snapshot hosts and has been manually applied to the copy of the repo in use for the current runs. Once the next job starts and completes for any wiki, the index.html file should be updated with the proper links for the combined multistream files.

If any large wikis are already in the last pass and won't pick this up, we can run a noop on those once their runs have completed, to fix things up.

Running noop jobs now for the big wikis that need it, in batches: arwiki dewiki commonswiki frwiki eswiki hewiki huwiki itwiki jawiki kowiki metawiki nlwiki plwiki ptwiki ruwiki

The above are done. Now running enwiki noop.

The noop for enwiki is done. A noop for wikidatawiki will be run once the April 1 dump is complete.

Change 589005 had a related patch set uploaded (by ArielGlenn; owner: ArielGlenn):
[operations/dumps@master] unit test for checking content of index.html file for a wiki dump run

https://gerrit.wikimedia.org/r/589005

Change 589005 merged by ArielGlenn:
[operations/dumps@master] unit test for checking content of index.html file for a wiki dump run

https://gerrit.wikimedia.org/r/589005

Closing this, as all links have been fixed up.