EOS DevOps Meeting

Tuesday 8 May 2018, 16:00 → 17:50 Europe/Zurich

513/R-068 (CERN)

Jan Iven (CERN)

Weekly meeting to discuss progress on EOS rollout.

• please keep content relevant to (most of) the audience, explain context
• Last week: major issues, preferably with ticket
• This week/planning: who, until when, needs what?

Add your input to the "contribution minutes" before the meeting. Else will be "AOB".

 

● EOS clients, FUSE(X)

(Dan) eos-fuse 4.2.22 needed to be hand submitted to the koji machin because of a hard dep on xrootd 4.8.3 (which is not yet in epel). So xrootd 4.8.3 has been added to the eos6/7 koji repo and 4.2.22 was rebuilt manually. Those rpms are now in qa. https://its.cern.ch/jira/browse/CRM-2662

Note that the i386 support for el6 has been dropped due to a missing software collections dependency. Will this be a problem?

 

 

(Andreas)

eosxd Long-term test on eosplusmey701 to PPS (after bug report from Massimo):

- created 1M files on mount - worked fine

- reading back 1M files from same mount and check checksums - in progress ( no error so far)

Development work (✔ = ready for merge)

EOS-2026 - [Andreas] - support small file inlining in eosxd ✔

EOS-2460 - [Rainer] - adds local FIFO support to eosxd ✔

EOS-2527 - [Rainer] - ACL management via richacl tools for eosxd ✔

EOS-2535 - [Andreas] - support temporary ETAGs for transparent service migration ✔ @eoshome

EOS-2551 - [Andreas] - several FSTs with individual HTTP ports on one physical machine ✔ @eoshome

EOS-2523 - [Andreas] - solved with EOS-2560 ✔

EOS-2552 - [Andreas] - added 'router' interface for XRootD & HTTP traffic to do directory to service redirection ✔ @eoshome

EOS-2554 - [Andreas] - do bulk recursive deletion on eosxd and move tree into recycle bin if configured - great performance boost ✔ @eoshome

EOS-2560 - [Andreas] - after discussion with Dirk adding also replica unlinking and namespace deletions to the report log file ✔

EOS-2563 - [Andreas] support several MGMs on one physical node with individual ports ✔ @eoshome

EOS-2564 - [Andreas] fix error.log log file population ✔

EOS-2565 - [Andreas] manage submounts in eosxd and squashfsfuse wip

EOS-2566 - [Andreas] support UID>1M in eos(x)d (if the UID range is exhausted only the user id is encoded and the server assigns the default group) ✔

EOS-2569 - [Andreas] fst:io:rados replacement plug-in (fst:io:exos) with EC support wip

 

 

16:00 → 16:02 (new meeting agenda)

16:00 → 16:20 EOS production instances (LHC, PUBLIC, USER)

• major events last week
• planned work this week

Speakers: Cristian Contescu (CERN), Herve Rousseau (CERN), Hugo Gonzalez Labrador (CERN), Luca Mascetti (CERN)

16:20 → 16:25 EOS clients, FUSE(X)

• (major) issues seen
• Rollout of new versions and FUSEX

Speakers: Dan van der Ster (CERN), Jan Iven (CERN)

(Dan) eos-fuse 4.2.22 needed to be hand submitted to the koji machin because of a hard dep on xrootd 4.8.3 (which is not yet in epel). So xrootd 4.8.3 has been added to the eos6/7 koji repo and 4.2.22 was rebuilt manually. Those rpms are now in qa. https://its.cern.ch/jira/browse/CRM-2662

Note that the i386 support for el6 has been dropped due to a missing software collections dependency. Will this be a problem?

 

 

(Andreas)

eosxd Long-term test on eosplusmey701 to PPS (after bug report from Massimo):

- created 1M files on mount - worked fine

- reading back 1M files from same mount and check checksums - in progress ( no error so far)

Development work (✔ = ready for merge)

EOS-2026 - [Andreas] - support small file inlining in eosxd ✔

EOS-2460 - [Rainer] - adds local FIFO support to eosxd ✔

EOS-2527 - [Rainer] - ACL management via richacl tools for eosxd ✔

EOS-2535 - [Andreas] - support temporary ETAGs for transparent service migration ✔ @eoshome

EOS-2551 - [Andreas] - several FSTs with individual HTTP ports on one physical machine ✔ @eoshome

EOS-2523 - [Andreas] - solved with EOS-2560 ✔

EOS-2552 - [Andreas] - added 'router' interface for XRootD & HTTP traffic to do directory to service redirection ✔ @eoshome

EOS-2554 - [Andreas] - do bulk recursive deletion on eosxd and move tree into recycle bin if configured - great performance boost ✔ @eoshome

EOS-2560 - [Andreas] - after discussion with Dirk adding also replica unlinking and namespace deletions to the report log file ✔

EOS-2563 - [Andreas] support several MGMs on one physical node with individual ports ✔ @eoshome

EOS-2564 - [Andreas] fix error.log log file population ✔

EOS-2565 - [Andreas] manage submounts in eosxd and squashfsfuse wip

EOS-2566 - [Andreas] support UID>1M in eos(x)d (if the UID range is exhausted only the user id is encoded and the server assigns the default group) ✔

EOS-2569 - [Andreas] fst:io:rados replacement plug-in (fst:io:exos) with EC support wip

 

 

16:25 → 16:35 Development issues

• New namespace
• Testing
• Xrootd

Speakers: Andreas Joachim Peters (CERN), Elvin Alin Sindrilaru (CERN), Georgios Bitzes (CERN), Jozsef Makai (CERN), Michal Kamil Simon (CERN)

16:35 → 16:50 AOB