Taskcluster/Meetings/2014-01-30: Difference between revisions
< Taskcluster | Meetings
Jump to navigation
Jump to search
(Created page with "== Previous Action Items == == Status Updates == === Job Queue (owner: dminor/ted) === === Provisioner (owner: jonas) === === Worker (owner: lightsofapollo, jonas, ted) ==...") |
m (Callek moved page TaskCluster/Meetings/2014-01-30 to Taskcluster/Meetings/2014-01-30: Pinged directly by :dustin to do this) |
||
| (9 intermediate revisions by 5 users not shown) | |||
| Line 4: | Line 4: | ||
=== Job Queue (owner: dminor/ted) === | === Job Queue (owner: dminor/ted) === | ||
* Modified Job message json to contain claim and finish endpoints. | |||
* Clean up/fixes to testing code. | |||
* Looking to merge in fixes for problems that show up while integrating worker and provisioner. | |||
* Next steps: Look at timing out RUNNING jobs, recovering jobs from dead workers. | |||
=== Provisioner (owner: jonas) === | === Provisioner (owner: jonas) === | ||
=== Worker (owner: lightsofapollo, jonas, ted) === | === Worker (owner: lightsofapollo, jonas, ted) === | ||
- initial version ready (no live logging) but conforms to the API we set down | |||
=== Task Graph (owner: aki) === | === Task Graph (owner: aki) === | ||
* Still not started yet due to b2g and mergeduty. I'll be at a dentist appt at this time; let me know if anyone has any questions for me. | |||
=== Treeherder Integration (owner: jeads) === | === Treeherder Integration (owner: jeads) === | ||
=== End-to-End Testing (ted) === | === End-to-End Testing (ted) === | ||
- CRAZY PUBLIC IPS EVEYRWHERE (ngrok) | |||
== Round Table == | == Round Table == | ||
* Think about interacting with [[ReleaseEngineering/Applications/RelengAPI]]? --dustin | * Think about interacting with [[ReleaseEngineering/Applications/RelengAPI]]? --dustin | ||
* Do existing tools like celery meet the need of the job queue? If not, what are they missing? --dustin | |||
* Suitability of http://paas.mozilla.org as a platform for services? --dustin (w/ cturra) | |||
* [dminor] What (if any) worker results should be passed to the job queue as part of calling finish? | |||
** Maybe when the job is claimed it should pick up a link the log and that is sufficient | |||
== Action Items == | == Action Items == | ||
Latest revision as of 18:38, 3 July 2017
Previous Action Items
Status Updates
Job Queue (owner: dminor/ted)
- Modified Job message json to contain claim and finish endpoints.
- Clean up/fixes to testing code.
- Looking to merge in fixes for problems that show up while integrating worker and provisioner.
- Next steps: Look at timing out RUNNING jobs, recovering jobs from dead workers.
Provisioner (owner: jonas)
Worker (owner: lightsofapollo, jonas, ted)
- initial version ready (no live logging) but conforms to the API we set down
Task Graph (owner: aki)
- Still not started yet due to b2g and mergeduty. I'll be at a dentist appt at this time; let me know if anyone has any questions for me.
Treeherder Integration (owner: jeads)
End-to-End Testing (ted)
- CRAZY PUBLIC IPS EVEYRWHERE (ngrok)
Round Table
- Think about interacting with ReleaseEngineering/Applications/RelengAPI? --dustin
- Do existing tools like celery meet the need of the job queue? If not, what are they missing? --dustin
- Suitability of http://paas.mozilla.org as a platform for services? --dustin (w/ cturra)
- [dminor] What (if any) worker results should be passed to the job queue as part of calling finish?
- Maybe when the job is claimed it should pick up a link the log and that is sufficient