Page MenuHomeMiraheze

[New] Server Resource Request for MediaWiki
Closed, ResolvedPublic

Description

Number of servers requested: 2
Service: mw*
Processor: 6
Memory: 5GB
Disk: 25GB SSD
Network: IPv6

Justification for request: With cloud14 down, we are running on only four MediaWiki servers which is making us extremely slow for the remaining wikis. So I am requesting two additional (temporary until we get cloud14 up again) MediaWiki servers on the remaining cloud servers, if we have the resources for it.

Endorsement by Engineering Manager (MediaWiki) or Site Reliability Engineer: @Reception123 (@John needs to approve as well)

Related Objects

Event Timeline

Unknown Object (User) created this task.Nov 17 2022, 17:08

Looking at P386 it looks like cloud12 or cloud13 could handle it.

This seems reasonable given to me in the circumstances but as mentioned above it would also have to be approved by John

Unknown Object (User) added a comment.Nov 17 2022, 19:42

This seems reasonable given to me in the circumstances but as mentioned above it would also have to be approved by John

Maybe another option is increasing resources on the remaining mw* servers for now also? It is hard to deploy new servers without puppet as far as I am aware.

In T9974#201243, @Universal_Omega wrote:

This seems reasonable given to me in the circumstances but as mentioned above it would also have to be approved by John

Maybe another option is increasing resources on the remaining mw* servers for now also? It is hard to deploy new servers without puppet as far as I am aware.

That's true, that's probably an easier option then

As I mentioned on IRC: How would we setup the new VMs without puppet141? (we can't get puppet running without the master) I guess you could do it completely manually.

In T9974#201243, @Universal_Omega wrote:

This seems reasonable given to me in the circumstances but as mentioned above it would also have to be approved by John

Maybe another option is increasing resources on the remaining mw* servers for now also? It is hard to deploy new servers without puppet as far as I am aware.

That's true, that's probably an easier option then

This seems like the more feasible option. Let’s decide what we want to change and then ask John to review.

Unknown Object (User) assigned this task to Reception123.Nov 18 2022, 01:53
Unknown Object (User) raised the priority of this task from Normal to High.
In T9974#201243, @Universal_Omega wrote:

This seems reasonable given to me in the circumstances but as mentioned above it would also have to be approved by John

Maybe another option is increasing resources on the remaining mw* servers for now also? It is hard to deploy new servers without puppet as far as I am aware.

That's true, that's probably an easier option then

This seems like the more feasible option. Let’s decide what we want to change and then ask John to review.

The main thing is probably increase each existing mw server +3 cores (makes up for the ones we lost keeping things balanced)

Per the latest update I believe this will no longer be needed

Unknown Object (User) reopened this task as Open.Nov 18 2022, 20:28
Unknown Object (User) closed this task as Resolved.Nov 19 2022, 00:14
Unknown Object (User) claimed this task.

@Reception123 approved me to increase test131 resources (at least temporarily) to match production as an emergency action. Test131 is now pooled into production and the increase of test131 resources might be enough for this for now to provide stability to users in the absence of mw14*

Just stating here for the record that under the policy that is an emergency approval by the EM MediaWiki due to the current situation with cloud14.