wiki:developers-meeting-20140129
Last modified 6 years ago Last modified on 01/29/14 16:29:22

[part of a series of meetings]

Participants

Dmitry, Al, Tigran, Christian, Paul

Agenda

[see box on the right-hand side]

Postcards

Up to two minutes (uninterrupted) per person where they can answer two questions:

  • What I did last week (since the last meeting),
  • What I plan to do in the next week.

No questions until we get through everyone :)

Al: finishing getLocality getChecksum code; bringOnline code work be testing. Problem with NFS v3 getting "permission denied", getSize, bringOnline didn't work. Dmitry&Al, entry into the replica manager changes; looking at the RM schema. Integrating replicamanager with chimera schema? Looked at a couple ... looking at Xavier ticket

Dmitry: worked a bit on quota -- Discussing with replica manager -- link to look at replica-manager ideas sent . looking at Gerd space-manger; trying to get encp to work with truck version of dCache -- discovered various NFS-related problems. Tested pcells and discovered the passwordless lkogin didn't work.

Paul: PoF prepartion, SRM delegation work.

Christian: pcells release created; looking into the ssh-keys problem just now. A couple of issues with SRM reported in tickets. One issue was that admin had the idea that srm-client would always use active transfers. Problem was configuration change. ... plus other tickets.

Tigran: working on a couple of NFS problems found at DESY. Made new version of NFS code; integrated into trunk. Updated 2.7 instance here at DESY with custom version. Rest of the time working on OnCloud?: have dCache with primary storage with NFS. Cristian's cloud the secondary source; synchronisation server --> client; primary copy it works. Sync. with the secondary copy isn't working at the moment. Patches sent upstream (some already accepted), to get OwnCloud? work out-of-box. Android/iPhone clients isn't tested yet, but that's next.

Plans for patch-releases

Should we make a new patch release?

Release everything.

In dcap, there is a patch dlopen problem. This may trigger another dcap release.

Trunk activity

Progress with new features...

Paul working on GridSite? delegation.

Replica Manager stuff

--> dynamic pool-group, add pools and remove pools

--> multiple groups

tons of scripts to monitor, scheduler inside dCache to do do something. Could we design replica-manager to use a more generic scheduling framework? This could be used for other activity, like gathering monitoring information.

Interesting idea; to be discussed further: possibly at a dev. workshop.

Try to fix replica-manager first.

Some discussion about replica-manager.

Dmitry to update wiki page based on discussion and send the link around team list again.

Issues from [FIXME: Add link to yesterday's Tier-1 meeting]

Xavier reported two problems:

problem #1 pool becomes time-out in admin interface when No. movers reaches 50. Pool unresponsive (have it here at DESY & Fermi). Waiting for thread-dump.

problem #2 many movers, many threads; can see only one CPU loaded out of ~20. Tigran suspects this is related to problem #1.

Anticipate 2.2 by end of February.

New or noteworthy

Approached by OSG security office cert. handling in dCache: Q. DOE don't want to support so many certificates "OSG production farm" work.

Outstanding RT Tickets

[This is an auto-generated item. Don't add items here directly]

Review of RB requests

DTNM

Same time, next week.