Last modified 3 years ago Last modified on 06/08/16 15:40:45

dCache Tier I meeting June 7, 2016

[part of a series of meetings]

Present, Paul), IN2P3(), Sara(), Triumf(), BNL(), NDGF(Ulf), PIC(Marc), KIT(Xavier), Fermi(), CERN()


(see box on the other side)

Site reports


Everything OK at PIC.

The only problem is with the NFS and space-manager

Tigran: it's a bit difficult to resolve this, because of how NFS handles retries. We need to discover where to update the code handle retries.

Kill the current transfer, recovers.

Tigran: Try creating another link, but only for NFS protocol and point this link to the pools that are also part of the space-token.

Discuss this during tomorrows Devel meeting.


On 2.16 for a while; upgraded, a little while ago (SNAPSHOT), now running 2.16.0.

So far, no problems.

Next week, upgrading PostgreSQL to 9.5. Tests show it should take about 5 minutes to do the update. Anticipate using HUGE_PAGES, to releaves memory fragmentation.

See INDIGO page for details of pre-GDB talk.


Things are running fine again.

Wednesday had a problem with the SRM database; controller failed. Unable to make the database run without replacing the controller and restoring from an earlier dump. Now things are running smoothly.

Xavier thanked Tigran for ability to recursively delete tags.

Running tests on PowerPC architecture. Tested with writing and reading via NFS. Xavier can test updated version of dCache scripts.

Xavier has started development effort that provides HPSS support for dCache, initially based on the fuse support in HPSS.

Currently we don't expose dCache API package; makes development hard.

Support tickets for discussion

[Items are added here automagically]


Same time, next week.