http://indico.cern.ch/conferenceDisplay.py?confId=46745
CERN: Steven, Oliver, Francesco, John, Maite.
Telephone: Jeff, (who else?) Vangelis, Jeroen.
SN: Minutes of last meeting... minor revisions from Christoph Witzig...
Anything else?
(no)
Action list:
8954: Will distribute new version. Oliver: will he add the URL to the task?
8953: Action still with Christoph. Jeff: It's in the document. WMS use case.
Oliver: does AuthZ understand different policies to different servers?
8164: General proxy store. Updated. Short term OK. Long term needed
8060: JSPG. We change the logic of the SAM tests to avoid the site being "bad".
Should close this now. ACTION.
7932: Later...
7841:
6901: Vangelis: Had a meeting with Tiziana Ferrari from INFN who is
responsible for the small team of INFN that is part of the NA4
Direct User Support team. Adding a note to the task that they (INFN)
produce a set of Data Management Use Cases that will be put on-line
together with the rest of the other UIG use cases produced in the
past (and whose revision is also responsibility of DUS).
Close 7841
Put a note in 6901... leave with Vangelis but note for INFN to do something.
7675: We have a doc. Close this?
JT: How about a standard interface?
OK: There is a new WG in OGF... CREAM/ARC/UNICORE.
JT: LCG-CE at least supports GRAM.
Close 7675. Nick Thackray to provide a link to the doc
Close 7106
6901: Done. Already discussed above.
6652: Characterizing a subcluster? Is there a proposal?
OK: Should be. Should be done properly in GLUE 1.3.
SN: Should this go into GLUE-2?
6649: OK: This could be useful for MPI. NA4 should make sure GLUE-2
supports this use-case.
Parked until GLUE-2 is out
(Long explanation from JT.)
SN: Is there a spec document?
FG: Twiki page but no document.
End of Action List
------------------
MPI WG:
------
Jeroen: (General info... recent history of the group, who joined etc)
(This list can be sent to me afterwards... 7 people)
(Also contacted Vincenzo Innocente @ CERN. No go)
SN: Goals, Milestones?
Jeroen: 23rd Feb Phone conference to divide work (what's the work?)
SN: Fusion has the biggest reqs for MPI.
If they cannot provide anyone to help... should be followed up.
SN: Please set up your goals and milstones.
Jeroen: Will do. Also a set up a FtF. Tried to use UF in Catania.
FG: Anyone from the MW community?
JT: Dennis van Dok is good guy for this.
SN/FG: Could try someone from WMS and also CREAM.
The INFN people should have replied to you when contacted.
Jeroen: Will do.
AuthZ Deployment Doc
--------------------
CW: There are slides sent out. These are attached to the agenda:
http://indico.cern.ch/conferenceDisplay.py?confId=46745
(Refer to the slides)
JT: How mature are these policy engines?
SN: We don't have to own the engine... we can use a mature
product from another project with no problems.Several under evaluation.
JT: Is there a tag on the PEP-client that gives the address?
Configured in client on WMS and also the AuthZ service
on the central service?
Could I do the same on my laptop and get the same info?
CW: Yes.
SN: Is the AuthZ service deployed at a site?
CW: Yes.
(Any other questions?)
CW: Slide 4. Global banning policy. See the slide 5.
SN: How close to doing a glexec and back test of the system?
CW/JW: Very close. Next two weeks.
SN: Any objections to the TMB endorsing this plan?
Are there any problems being seen
Any PPS sites foreseen for AuthZ system?
JT: How about code re-use from SCAS to EES?
CW: We don't know but the plugin system will allow
much re-use of "SCAS".
Errors in gLite
---------------
SN: Francesco G. has indicated this is a "null" activity.
FG: Indeed.
OK: All our testing is based on replication of real-life situations.
JT: This type of thing was discussed in the MB recently.
Developers fixed a problem that was not really a problem.
This is due to not really having specification docs on APIs etc.
SN: This is true for UMB etc. The MU of EGI will have this problem.
Need a preparation of documents of all public interfaces
before the end of EGEE-III.
(list of bugs that Steven has... these bugs are probably not going to advance)
FG: This will not be easy to fit into our work plan now.
People are all booked up.
AOB
---
SVN question: Developers would like to be able to use repositories other than the standard CERN CVS. Other systems are reputed to have better behaviours for developer-specific tasks.
SN (TMB in general): TMB will not restrict use of svn. Any software repository should be institutionally supported. ie. No bizarre systems hosted on laptops!
There are minutes attached to this event.
Show them.