We classify "large" wikis as ones with more than 300k entities.
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | Goal | Martyn.ranyard | T368356 ⬆️Update Wikibase Cloud to MediaWiki 1.43 | ||
| Resolved | dena | T411637 ⬆️ N3: Update remaining "large" wikis to MW1.43 on production |
Event Timeline
Comment Actions
As of now these are:
lexbib.elex.is osloddt.wikibase.cloud biodiversity.wikibase.cloud wikifcd.wikibase.cloud geokb.wikibase.cloud addshore-wikibase-cloud-status.wikibase.cloud domains.wikibase.cloud social-contagion.wikibase.cloud
where as addshore-wikibase-cloud-status.wikibase.cloud was already updated T411634: ⬆️ Update Addshore's status wiki
Comment Actions
Together with @dang I kicked off k8s jobs for batch N3 wikis at Fri, 12 Dec 2025 14:12:04 +0100
- first with lexbib.elex.is
- and shortly after for
osloddt.wikibase.cloud biodiversity.wikibase.cloud wikifcd.wikibase.cloud geokb.wikibase.cloud domains.wikibase.cloud social-contagion.wikibase.cloud
- Same approach was followed as for T411635: ⬆️ N2: Update remaining "small" wikis to MW1.43 on production
- running a temporary script, first like
temp_n3.sh
#!/bin/bash set -euo pipefail WIKIS=" lexbib.elex.is " for WBS_DOMAIN in $WIKIS; do echo "next up: $WBS_DOMAIN" WBS_DOMAIN=$WBS_DOMAIN MW_VERSION=mw1.43-wbs1 MEDIAWIKI_BACKEND_INSTANCE_LABEL=mediawiki-143 ./mediawikiUpdate.sh done exit osloddt.wikibase.cloud biodiversity.wikibase.cloud wikifcd.wikibase.cloud geokb.wikibase.cloud domains.wikibase.cloud social-contagion.wikibase.cloud
and then after seeing it successfully starting off the first job, re-running it like this:
temp_n3.sh
#!/bin/bash set -euo pipefail WIKIS=" osloddt.wikibase.cloud biodiversity.wikibase.cloud wikifcd.wikibase.cloud geokb.wikibase.cloud domains.wikibase.cloud social-contagion.wikibase.cloud " for WBS_DOMAIN in $WIKIS; do echo "next up: $WBS_DOMAIN" WBS_DOMAIN=$WBS_DOMAIN MW_VERSION=mw1.43-wbs1 MEDIAWIKI_BACKEND_INSTANCE_LABEL=mediawiki-143 ./mediawikiUpdate.sh done exit
So far, as expected, we still see warnings in the logs like this:
[rdbms] Expectation (writeQueryTime <= 5) by MediaWiki\\Maintenance\\MaintenanceRunner::run not met (actual: 8.3101811408997)
Comment Actions
Completed after less than 24 hours!
$ date && ./verifyMediawikiUpdates.sh Sat Dec 13 01:53:11 PM CET 2025 Completed jobs: mediawiki-update-2ml84 osloddt.wikibase.cloud mediawiki-update-4jtqz biodiversity.wikibase.cloud mediawiki-update-65bzj geokb.wikibase.cloud mediawiki-update-bpfrn domains.wikibase.cloud mediawiki-update-tk2xc lexbib.elex.is mediawiki-update-vjwx6 wikifcd.wikibase.cloud mediawiki-update-w4fnv social-contagion.wikibase.cloud Incomplete jobs: --- Number of completed jobs: 7 Number of incomplete jobs: 0
We got a couple of new errors in the alerting/error reporting, but these were all still from MW 1.39 pods during the updates.