411335 - Sorroco server changes that block next beta and final release of firefox 3.

Status: RESOLVED FIXED

(Socorro :: General, task)

Description

[chris hofmann]

most of these problems are currently described over in http://code.google.com/p/socorro/issues/list

most critical are the need for comment analysis, access to raw stack data and registers, and any bugs that make performance of the server to slow to do timely analysis.

At this time we know we need a server drop or mozilla.org infra changes to make sure this set of problems are resolved
 
aggregate display of comments for topcrashers
http://code.google.com/p/socorro/issues/detail?id=117

Store/Display raw stack data and registers
http://code.google.com/p/socorro/issues/detail?id=118

getting to top crash reports is really slow.
http://code.google.com/p/socorro/issues/detail?id=119

Comment 1

[chris hofmann]

since there are no blocking flags on server site components I did the ulgy think of changing the product to firefox general so these set of bugs come up on the appropriate discussions and stop-ship lists...

Comment 2

[Benjamin Smedberg]

I do not think that raw stack data/registers should be a blocker for anything... we have a symbol server now for releases, so the only case it would be really useful is a bug that is entirely unreproducable by our devs and larger testing community but is very common in the real world, which seems like a very uncommon case to me.

Comment 3

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

(In reply to comment #2)
> I do not think that raw stack data/registers should be a blocker for
> anything... we have a symbol server now for releases, so the only case it would
> be really useful is a bug that is entirely unreproducable by our devs and
> larger testing community but is very common in the real world, which seems like
> a very uncommon case to me.

That's been quite common in the past -- generally because of bugs that occur randomly, at very low probability, for common steps to reproduce (e.g., missing short-term GC roots).

Comment 4

[Benjamin Smedberg]

I think the difference is previously, savvy users in the wild (or even moz developers) couldn't debug such crashes in the field. Now with the symbol server, they can.

Comment 5

[David Baron :dbaron: (⌚️UTC-5, no longer working on Mozilla)]

But somebody hitting one of the top random crashes doesn't necessarily know that that's what it is -- and how would they know until after they've submitted the report?

Comment 6

[Samuel Sidler (old account; do not CC)]

The issues listed in comment 0 are filed as bug 411430, bug 411431, and bug 411432, respectively.

Comment 7

[Mike Schroepfer]

Moved all bugs in comment 6 to blocking list - moving this bug off...

Comment 8

[Michael Morgan [:morgamic]]

No need for this bug, going off the flags now -- resolving.