Page MenuHomePhabricator

Use Sentry in production
Closed, DeclinedPublic

Description

Noticing, tracking and debugging errors should be easy. A Wikimedia developers and users should be able to easily notice new errors (possibly only a subset of them that they are interested in), find out details, and share with others. At the same time security and privacy standards should be enforced.

https://getsentry.com/

Related Objects

StatusSubtypeAssignedTask
DeclinedNone
OpenNone
DeclinedNone
DeclinedTgr
ResolvedTgr
ResolvedTgr
Resolvedcsteipp
ResolvedTgr
ResolvedTgr
ResolvedAklapper
ResolvedTgr
ResolvedTgr
OpenNone
ResolvedTgr
DeclinedTgr
DeclinedTgr
DeclinedTgr
ResolvedTgr
DeclinedTgr
ResolvedTgr
ResolvedTgr
ResolvedKrinkle
DeclinedNone
OpenNone
ResolvedTgr
DeclinedNone
OpenNone
InvalidNone
DeclinedTgr
ResolvedTgr
DeclinedNone
DeclinedNone
DeclinedNone
ResolvedTgr
Resolvedjcrespo
DeclinedTgr
OpenNone
OpenNone
ResolvedJdlrobson
ResolvedJdlrobson
ResolvedJdlrobson
ResolvedJdlrobson
ResolvedNone
Resolvednray
ResolvedJdlrobson

Event Timeline

Tgr raised the priority of this task from to Low.
Tgr updated the task description. (Show Details)
Tgr added projects: Sentry, Epic.
Tgr added a subscriber: Tgr.
Jdlrobson added a subscriber: Jdlrobson.

Would be great to get buy in from the frontend standards group to move this along!

Being bold and adding the Developer-Wishlist tag.

If I understand correctly what this technology does, deploying it will help me a lot, both as a developer of Gadgets and as a developer of MediaWiki features and extensions that are deployed on a lot of sites and clash with local gadgets much more frequently than anybody would like to.

Unfortunately this is probably a bit beyond the scope of what "will take a developer in their volunteer time a few days, or up to a month". For one thing, it would require dedicated hardware. T91649: Deploy Sentry (JavaScript error logging) to production, configured to log only a limited subset of users/pages or something along those lines (maybe limited to logged-in users instead of limited to a single page) might be more realistic.

For what is worth, I believe that the Developer-Wishlist (2017) can be used to grasp the level of interest of an idea. We told ourselves that we would allow some experimentation.

The title for the survey could be "Agree on a plan to use Sentry in production". The possible plans agreed could be many, included "Not in the next 18 months", but at least there would be a discussion in the good direction.

This can be descoped and still be meaningful. I think I'll replace it with something like "log Javascript errors for logged-in users on medium and small wikis" (huge wikis usually have better mechanisms to notice bugs anyway).

This proposal is selected for the Developer-Wishlist voting round and will be added to a MediaWiki page very soon. To the subscribers, or proposer of this task: please help modify the task description: add a brief summary (10-12 lines) of the problem that this proposal raises, topics discussed in the comments, and a proposed solution (if there is any yet). Remember to add a header with a title "Description," to your content. Please do so before February 5th, 12:00 pm UTC.

As noted in T91649 (duplicating here just for more dissemination), sentry is no longer open source (by the strict definition of the term). More at https://forum.sentry.io/t/re-licensing-sentry-faq-dAiscussion/8044

Can we drop deployment-sentry01 from beta cluster? Beta cluster has reached its quota and I'm trying to clean it up a bit.

Can we drop deployment-sentry01 from beta cluster? Beta cluster has reached its quota and I'm trying to clean it up a bit.

Yes please. I don't think it works, and in any case, plans of deploying Sentry to production have been abandoned, and even if we were to pick them up again, Sentry has changed too much for the version deployed in beta to be of any use.

Mentioned in SAL (#wikimedia-releng) [2020-07-04T23:51:27Z] <Amir1> deleted deployment-sentry01 (T106915)

Peachey88 added a subscriber: Peachey88.

plans of deploying Sentry to production have been abandoned.

Boldly declining this task for now.

(FWIW, Sentry was abandoned because 1) the main use case was JS error logging and there T226986 captures most of the value with much less effort, 2) its licence is not OSI compatible anymore, 3) the way it is packaged would have made it hard to use it in production. Extension:Sentry, which can connect to the SaaS, is still meaningful for third parties, and there is still some interest in using Sentry, or its FLOSS fork, in Wikimedia Cloud - that's tracked in T106913: Use Sentry on non-production Wikimedia wikis, Toolforge and other sites/tools.)