Cryptograss:Emergents/Magent/Tasks/Jenkins red: Difference between revisions
From PickiPedia: A knowledge base of bluegrass, old time psychedelic jams, and other public domain music
Jump to navigationJump to search
Draft task: wake magent in #ops when a Jenkins check stays red (magenta#402) (via create-page on MediaWiki MCP Server) |
m Verified bot edits (removed proposal markers) |
||
| Line 1: | Line 1: | ||
{{Emergent task|mood=ops|cap=3|tools=look}} | {{Emergent task|mood=ops|cap=3|tools=look}} | ||
<!-- Drafted by magent, 5 Nov 2026 (#ops, Justin: "red runs oughta hit the Task API and wake you in here"). Started by magenta (services/alarms.py, still_red) when a Jenkins check has stayed red past one cron-health round; never on a schedule. Runs only once JMyles or SkymanJenkins saves a revision. --> Jenkins on maybelle says '''{job} · {stage}''' has been red since {since} and didn't clear on the next run. The build: {url} | |||
* | * Find out why, from what you can read: the stage's script in maybelle-config's <code>maybelle/jobs/{job}.groovy</code> at production, the deploy logic it checks, and the recent record here (redeploys, earlier alarms). Say what you checked. | ||
* | * If a person has to look on maybelle (a log, a file the stage reads), give them the one read-only command that settles it. | ||
* | * If the fix is in deploy logic, say so and ask for tools to open the PR. Don't fix a server by hand. | ||
* | * If it's the check that's wrong (a threshold too tight, a flap), say that, and what you'd change. | ||
Keep it short: one post, and mention @justin only if a person has to act. | |||
Latest revision as of 03:46, 7 November 2026
A task for Magent, run by memory-lane when its schedule comes round or a person starts it.
- Mood: #ops
- Schedule: none: only when a person starts it
- Cap: $3 a day in that Mood
- Tools: look-only
Only a revision by a trusted editor (JMyles, SkymanJenkins) counts. Everything on this page after this box, less <!-- comments -->, is what the emergent is sent; a {name} is a blank filled in when it's started.
Jenkins on maybelle says {job} · {stage} has been red since {since} and didn't clear on the next run. The build: {url}
- Find out why, from what you can read: the stage's script in maybelle-config's
maybelle/jobs/{job}.groovyat production, the deploy logic it checks, and the recent record here (redeploys, earlier alarms). Say what you checked. - If a person has to look on maybelle (a log, a file the stage reads), give them the one read-only command that settles it.
- If the fix is in deploy logic, say so and ask for tools to open the PR. Don't fix a server by hand.
- If it's the check that's wrong (a threshold too tight, a flap), say that, and what you'd change.
Keep it short: one post, and mention @justin only if a person has to act.