Skip to content

mozilla-necko/necko-triage

 
 

Repository files navigation

necko-triage

A simple webapp to help sort things out in necko triaging

triage guide

Triage and Necko's bug handling pipeline

The purpose of Necko's triage process is to organize "new" networking bugs into their appropriate place for handling. Not all bugs are "new"; some come from other components, some are being revisited or re-opened.

In general, we want to:

  • Validate that bugs belong in the networking component (which are our responsibility)
  • Ensure that we have sufficient information to assign a priority and severity.
  • Categorize them under appropriate meta bugs for easy tracking
  • Alert performance team of performance impairments

There are two locations we watch for incoming bugs:

We use Bugzilla to edit the bugs.

Post-triage, a typical bug MAY flow through our handling stages like so:
Triage -> New -> Review -> Next -> Priority

We also have bugs that we are Monitoring closely that may require further action.

Also, If you’re filing a new bug you’ll be working on (or as part of a project you’re working on, even if you won’t be working directly on it), please triage appropriately, so as to not overload whoever is on triage that week :-D

Some quick links to Bugzilla

How to Triage

Add to team's upcoming priorities if it seems like something that should be handled soon

  • Add [necko-priority-review] to the "whiteboard" field to submit for bug review.
  • Emergency items can be put directly into [necko-priority-queue]
  • (See below Post-triage for more details)

Add to relevant meta bugs

  • Add the meta bug (ID or alias) to the "blocks" field of the new bug.
  • (See the triage helper for the list of existing meta bugs and their aliases)

Handle performance bugs

  • All performance related bugs should be added to necko-perf metabug.
  • If the issue is one that hurts performance (or you're not sure) set the "Performance Impact" flag to ?.

Request more information (when necessary)

  • Steps to reproduce. You may also want to confirm they actually reproduce.
  • http logs should be submitted to [email protected] to prevent leaking personal information.
  • about:crashes - submit crash data to crash-stats.mozilla.org
  • about:support - application and system details

Mark as triaged

  • Set [necko-triaged] in the whiteboard for bugs that have been triaged.
  • Set Priority (see below for more)
  • Set Severity

Set Rank/Points

  • Rank is an estimate on the impact of a bug, it notes the relative importance of P2/S3's against each other. Where 0 is the highest impact.
  • Points is an estimate of the complexity/difficulty of fixing a bug, where 1 is low complexity/easy.

Notes on Priorities

Priorities

  • P1 bugs (Blocker: get someone assigned ASAP: also EMAIL necko@moz whenever we have a new P1 bug)
  • P2 bugs (Active--we intend to work on these now or this quarter)
  • P3 bugs (backlog)
  • There is no P4--don't use it
  • P5 bugs (would take)

Principles/Rules of Thumb

  • P1 bugs MUST have a person assigned
  • Sec-highs are generally at least S2s, even with low crash frequency
  • S2s should be confirmed as real bugs before being labelled as such
  • Non-necko people can be assigned to necko bugs just fine
  • Bugs filed by wpt-bot should be marked triaged as P5. If something is needed there, someone can ni? one of us.

Post-triage

  • New is where we first put items that will be reviewed in the regular bug meeting. It helps readability of the newest items if too many Review items linger. After first viewing in the regular bug meeting a New item should be moved into Review.
  • Review is the bucket of bugs that get reviewed in our regular bug review meeting. Selected bugs will may get promoted for action. Review bugs should have a whiteboard with [necko-priority-review].
  • Next is essentially used to label bugs that have already been reviewed and agreed upon as going into the priority queue soon. This is useful when the review queue gets a bit bloated. Next bugs should have a whiteboard with [necko-priority-review] and [necko-next].
  • Priority is the bucket of bugs that are a team priority. We try to pull our bug fixing efforts from this list and keep it at a limited size. Priority bugs should have whiteboard [necko-priority-queue] label.
  • Monitoring is the bucket of bugs that we are closely watching. For example, we may be watching for crash spikes or whether a submitted patch has fixed a particular issue. These are relatively passive bugs that should be moved back into the priority pipeline if necessary. Use the whiteboard label [necko-monitor] for this bucket.

About

A simple webapp to help sort things out in necko triaging

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • JavaScript 86.1%
  • HTML 9.4%
  • CSS 4.5%