diff --git a/docs/meetings/2024/TechArea20241119.md b/docs/meetings/2024/TechArea20241119.md index 50ad7dd5e..188941cd6 100644 --- a/docs/meetings/2024/TechArea20241119.md +++ b/docs/meetings/2024/TechArea20241119.md @@ -2,7 +2,7 @@ - **Coordinates:** Conference: +1-415-655-0002, PIN: 146 266 9392, (password sent separately) -- **Attending:** Mat, Matt, Tim +- **Attending:** Marco, Mat, Matt, Tim ## Announcements @@ -17,54 +17,48 @@ Triage duty shifts Tue-Mon - 8 (+0) open FreshDesk tickets (since 11/05) - 0 (+0) open GGUS ticket -### Jira (as of Monday morning) (TODO) +### Jira (as of Monday morning) + +(delta since 11/05 meeting) | # of tickets | Δ | State | |--------------|---------|-------------------| -| 220 | -1 | Open | -| 17 | -4 | Selected for Dev | -| 45 | +1 | In Progress | +| 221 | +0 | Open | +| 17 | +0 | Selected for Dev | +| 45 | +0 | In Progress | | 17 | +0 | Dev Complete | -| 2 | +1 | Ready for Testing | -| 1 | +1 | Ready for Release | +| 1 | -1 | Ready for Testing | +| 0 | -1 | Ready for Release | -## OSG Software Team (TODO) +## OSG Software Team OSG 24 clean-up: - OSDF caches / origins - AI (Mat): prepare `osdf-server` for release with Pelican 7.11 - - AI (Mat): write cache doc, these upstream changes will be relevant: - - - AI (BrianL) write origin docs -- Container images: - - AI (Matt): upgrade instructions for the OSPool EP - - Following repos need OSG 24 added: - - `docker-compute-entrypoint` - - `docker-xcache` - - `gratia-probe` - - `htcondor-autoscale-manager` - - `open-science-pool-registry` - - `osg-repo-scripts` (?) - AI (Matt): Sync 24-main tarballs onto repo + - Ask UW CSL to set up rsync server +- AI (Matt): Fix various OSG 24 image build issues +- AI (BrianL): Review PRs for: + - OSG 23 to 24 upgrade docs + - HTCondor-CE 23 to 24 upgrade docs -EL9 repo: +EL9 repo: - AI (Mat): Write repo migration instructions ### Discussion -None this week +- GlideinWMS: 3.10.8 in progress +- LIGO uses startd disk enforcement, which is causing Glideins to request insufficient disk + and exit without running any jobs. + Marco will talk to James Clark to debug. -### Support Update (TODO) +### Support Update -- BNL (BrianL): Belle II SSL-based glideins aren't getting reported. - Not a supported method in OSG so we have a hacky workaround where they set `orig_AuthTokenIssuer` and GRACC will map - the value to Belle II. -- ATLAS / BNL / AGLT2 (BrianL, MattW): questions about scheduling prioritizing cores and memory efficiency. - Greg is taking point on the ATLAS communication thread, seems relevant for us. - Notably UChicago doesn't have these problems: they allow Harvester to specify the memory that they want and allow - pilots to consume 3x their memory requests +- Matt (Harrisburg): ospool-ep container not cleaning up scratch directories. + Mats Rynge is working on a cleanup script for factory-sent pilots; + Matt will investigate making use of that script for container pilots as well. ## DevOps