Crash reporting all hands

[Gabriele Svelto]

Hello all,
during our in-person all hands it had become a tradition to have a one
hour crash reporting roundtable with everybody interested in the topic.
Since we haven't had one in a while and since we've covered a lot of
ground in the past 18 months it might be worth talking face-to-face
again; even if it has to happen on Zoom.

Here's a short list of topics we might want to cover:

* Symbolication improvements (inlined functions, native debuginfo)
* Improvements around hang reporting (ability to capture hangs,
discoverability and bucketing)
* Ongoing crash reporter oxidation
* Increased role for client-side crash analysis (more extensive use of
crash pings, better stack traces in crash pings, etc...)
* Capturing early crashes
* Helpful crash reporting (e.g. automatically pointing users
experiencing crashes to known bugzilla or sumo pages)

Additionally the nightly crash triage roster is still not full, so it
would be nice to discuss how we want to deal with crash triage in
general, and how to make it fit within all our other activities.

I'm sure there's more so please send all your suggestions for topics
that might be worth discussing, especially the ones that are cross-org
and might involve and impact multiple teams.

I haven't thought of a date yet, maybe something after mid-October?

Gabriele

[Gabriele Svelto]

On 27/09/21 16:52, Chris H-C wrote:
> Count me in as a consultant. With upcoming changes to Firefox Telemetry,
> "crash" pings are (eventually) going to change. It might be worth
> spending some dedicated time on an Agenda talking about how crash
> reporting looks in Glean.

Thanks for reminding me, I had completely forgotten but it's a very
important topic. The transition from traditional crash pings to Glean is
definitely something we want to talk about.

Gabriele

[William Kahn-Greene]

I miss these meetings because they were a good way of figuring out needs and plans for the next six months. Also, it helped to figure out who was who. 

I’d like to get a status on all the projects in the air and which ones are active and which are on hold.

Socorro and Tecken are going to GCP in the next year which potentially gives us some opportunities for tech overhauls. Also, crash ingestion is part of Data Org and the future brings consolidation between those two worlds in every which way that helps. I’d be good to toss around what that entails.

Also, I’ve been talking about symbolication of crash pings for 5 years now. I’m wondering whether I am Captain Ahab and this is my whale. Changing the structure of crash pings could be really helpful.

Anyhow, I like this idea. We should keep notes and send a summary to these lists. Maybea blog post, too. I can own that. 

/will

--
You received this message because you are subscribed to the Google Groups "stab...@mozilla.org" group.
To unsubscribe from this group and stop receiving emails from it, send an email to stability+...@mozilla.org.
To view this discussion on the web visit https://groups.google.com/a/mozilla.org/d/msgid/stability/b3ec6056-78f5-37c1-a3ed-0263ab2530a3%40mozilla.com.