Changes between Version 1 and Version 2 of developers-meeting-20140925


Ignore:
Timestamp:
09/25/14 16:49:06 (6 years ago)
Author:
paul
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • developers-meeting-20140925

    v1 v2  
    55== Present == 
    66 
    7 dCache.org(), IN2P3(), Sara(), Triumf(), BNL(), NDGF(), PIC(), KIT(), Fermi(), CERN() 
     7dCache.org(Tigran, Paul), Triumf(Simon), KIT(Xavier), Fermi(Natalia) 
    88 
    99= Agenda = 
     
    1515== Fermi == 
    1616 
    17 NFS: downgrade 4.0 works better than 4.1. 
     17Natalia reported that, having reported problems mounting their dCache v2.2 with NFS v4.1, they tried mounting the NFS door using v4.0.  This worked better, so they will continue with this mount configuration. 
    1818 
    19 Pools usually have 3 disks usually attached; It's a long process when the hardware is replace or "upgrade".  If Fermi reboot the pool then they don't want the pool to start. 
     19At Fermilab, pool nodes normally have 3 disk systems attached to them.  A disk system may be upgraded or some other intervention without restarting the node, but the system group suggest restarting the whole node "just to be sure". 
    2020 
    21 Special role in puppet. 
     21Paul said that, if the OS is not affected by the intervention then the following procedure should be OK: stop the domain hosting the affecting pool; make the disk intervention; start the domain hosting the pool.  Rebooting the pool node is not a requirement from dCache. 
    2222 
    23 Reboot every-time  
     23Natalia also asked about the dCache workshop: when and where will the next one be held?  The paper-trail for travel must be started a minimum of two months before travelling and ideally three months.  The approximate date is enough: the dates can be changed later on.  It wasn't clear how approximate the location needs to be (is "in Germany" sufficient?). 
    2424 
    25 Need to know an approximate date and location for the dCache workshop 
     25Paul will ask within the team and try to discover how this will be handled. 
    2626 
    2727== KIT == 
    2828 
    29 dCache version 2.6.33  for LHCb xrootd door. 
     29Xavier joined the meeting to ask about the xrootd door problem.  He said that, in the ticket, Phillippe described the problem as a show-stopper; however, as the ticket existed for ~3 weeks, it isn't clear what is the real priority. 
     30 
     31Paul will investigate -- and ask Gerd if he has any idea. 
     32 
     33The LHCb xrootd door is running dCache version 2.6.33. 
    3034 
    3135== Triumf == 
    3236 
    33 Everything looks good so far. 
     37Simon is back from vacation: everything looks good so far. 
    3438 
    3539 
     
    4044= DTNM = 
    4145 
    42 Proposed: same time, next week. 
     46Tuesday and Thursday next week: Tigran will chair the meetings.