Attending: Jens (c+m), Marcus, Luke, John B, John H, Steve, RobC, Daniel, Brian, Sam, Govind, Winnie, Elena


0. Operational blog posts

   In particular the SNO+ space token, as discussed last week.  Snoplus seem to be both able and happy to use a space
   token with sites, so it has been proposed/agreed that they use SNODATADISK.  How should this be added retroactively?
   With DPM, it is thought the safest way is to set up a new area and copy over; Liverpool allocated 20TB.

1. Information systems and accounting and what's that vapor thing. I am
   still confused.

   While we, or I, are/am still somewhat confused, it may make sense to create a wiki page which attempts to
   summarise the current set of requirements for information publishing for all our stakeholders.

   For example, there is the ATLAS file upload, the EGI requirements, APEL, and the BDII monitoring formerly known as
   gstat, there are GLUE2 requirements and a long standing ticket on the T1 to publish GLUE2 for CASTOR, and there
   is whatever vapor is which is still not entirely clear to all of us but at least Luke found the documentation
   (aka source code).

   There is a data steering group for WLCG which could also be consulted.

2. KeyDocs - clearly made some progress, that's good. However, two
   important KeyDocs are left behind, Disk Pool Manager and StoRM.

   We agreed that Sam should own the wahidian documents for now; however, given that both Sam and Brian maintain
   important storage docs, we should ensure that the load is distributed equitably.

3. Whatever happened to that Ancient ATLAS Testing HammerCloud Thing?

   So.  The aim was to automatically test file access across sites; but the current Hammer cloud is not designed
   to operate in this way, as it will rather naturally always prefer local data.  Instead some relevant testing
   is done with CMS.

4. Events coming up - ISGC (it's early this year), Networkshop (= IPv6).
   There's also a LUG coming up but it's on the other side of the Pond
   (http://opensfs.org/lug-2017/).

   DavidC is going to ISGC, albeit with his security hat.  RobA submitting work on accounting for T1 storage resources.

5. ZFS - updates

   Marcus leaves at the end of March - so we lose the expertise thanks to Brexit - instead both Marcus and his
   expertise will head off to Victoria, Canada.

6. AOB

   Jens should notify Steve if/when the documentation mentioned in 1 gets created.


Daniel Traynor: (22/02/2017 10:01:57)
whos the sno+ contact
John Bland: (10:03 AM)
David Auty
Daniel Traynor: (10:05 AM)
I have added it to the sno+ space at qm hopefully it "just" works but some one needs to check 
John Bland: (10:05 AM)
auty@ualberta.ca
Brian Davies @RAL-LCG2: (10:05 AM)
https://www.gridpp.ac.uk/wiki/GridPP_VO_Incubator
Paige Winslowe Lacesso: (10:13 AM)
and for hdfs/dmlite?
Jens Jensen: (10:28 AM)
http://indico4.twgrid.org/indico/event/2/overview
https://www.jisc.ac.uk/networkshop
John Bland: (10:30 AM)
dan: the SNODATADISK token should have snoplus.snolab.ca/Role=production, yours just has snoplus.snolab.ca as far as I can see
Daniel Peter Traynor: (10:31 AM)
there was a good talk at the europen version of lug last year that made me aware of SAS switches. https://www.eofs.eu/_media/events/lad16/07_thiell_cheap_n_deep.pdf
Paige Winslowe Lacesso: (10:37 AM)
THANKS