Page MenuHomePhabricator

Provide dynamic report of differences between replica databases and production databases
Closed, ResolvedPublic

Description

Related to bug 54847.

I'd like the ability to regularly read a report that shows the differences between the Labs database replicas and the production databases. This would be helpful for:

  • finding new core or extension table columns or databases that have been deployed that aren't yet exposed via a view; and
  • analyzing what can and can't be additionally exposed (for example, watchlist and archive).

Basically this is a request for a regularly updated status report (or diff) between the two data sets. What's available on production vs. what's available on Labs.


Version: unspecified
Severity: normal

Details

Reference
bz55455

Related Objects

View Standalone Graph
This task is connected to more than 200 other tasks. Only direct parents and subtasks are shown here. Use View Standalone Graph to show more of the graph.

Event Timeline

bzimport raised the priority of this task from to Needs Triage.Nov 22 2014, 2:15 AM
bzimport set Reference to bz55455.
bzimport added a subscriber: Unknown Object (MLST).

Providing schema-only dumps of the databases from the Labs side and the production would allow for nice-ish diffs, maybe.

rodrigosprimo wrote:

I think that a good start would be to add a list of the missing tables and columns to the page below.

https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Database_access

chasemp subscribed.
jcrespo subscribed.

This will be eventually fixed with the parent task at the same time- this ticket will stay for the scope of making a part if it public.

LSobanski raised the priority of this task from Lowest to Low.Apr 26 2021, 10:09 AM
Marostegui subscribed.

I am going to close this as fixed. We've fixed major differences in schema and the new clouddb* hosts were cloned directly from production. We also run ROW based replication there which should ensure consistency on the data (otherwise replication will break).
As mentioned at T57455#2796252 once we've fixed the parent task that means we've fixed the differences and drifts present on the cloudd* hosts.

If someone believe this still needs to stay open, please reopen it.