Hi All...
We are living a transition phase from gLite 3.1 (SL4 i386) to glite 3.2
(SL5 s86_64). I have two question on how to best migrate and operate
with the resources:
### Issue 1 ###
To cope with the project request to migrate the resources to gLite 3.2
until September 2009, we are setting up different queues to support
those two different kinds of WNs. As an example, I have configured in
the same CE the dteam support for 2 different queues. If in my JDL I
include a requirement to my CE, the result of a list match will be:
-bash-3.00$ glite-wms-job-list-match -a test.jdl
Connecting to the service https://wms01.lip.pt:7443/glite_wms_wmproxy_server
=========================================================================
COMPUTING ELEMENT IDs LIST
The following CE(s) matching your job requirements have been found:
*CEId*
- ce02.lip.pt:2119/jobmanager-lcgsge-dteamgrid
- ce02.lip.pt:2119/jobmanager-lcgsge-dteamgrid_x86_64
=========================================================================
This means that the two queues are somehow ranked by the WMS, but both
queues offer resources, and a normal job may run in one queue or in another.
### Question 1 ###
In this situation, is there a way for a user to submit to a specific
queue without having to explicitly name it?
Starting from the example above, I would like to submit to the 64 bit
queue using a requirement from the information system. Nevertheless, I
only find "GlueHostArchitecturePlatformType" to determine the
architecture, and its content is an unique value per CE (in my case,
i686). Do I have to set a different CE for this ?
### Issue 2 ###
The information system is presenting a "GlueCEStateFreeJobSlots" value
(in the "dn: GlueVOViewLocalID" fields) representing the sum of
resources for both queues, when the "dn: GlueVOViewLocalID" field is
presented per queue. Check the info bellow. At this moment, tne number
of free job slots is 144 for "dteamgrid_x86_64" and 71 for "dteamgrid".
The following information reports 215 for each queue.
dn:
GlueVOViewLocalID=dteam,GlueCEUniqueID=ce02.lip.pt:2119/jobmanager-lcgsge-
dteamgrid,Mds-Vo-name=resource,o=grid
objectClass: GlueCETop
objectClass: GlueVOView
objectClass: GlueCEInfo
objectClass: GlueCEState
objectClass: GlueCEAccessControlBase
objectClass: GlueCEPolicy
objectClass: GlueKey
objectClass: GlueSchemaVersion
GlueVOViewLocalID: dteam
GlueCEAccessControlBaseRule: VO:dteam
GlueCEStateRunningJobs: 0
GlueCEStateWaitingJobs: 0
GlueCEStateTotalJobs: 0
GlueCEStateFreeJobSlots: 215
GlueCEStateEstimatedResponseTime: 0
GlueCEStateWorstResponseTime: 0
GlueCEInfoDefaultSE: se05.lip.pt
GlueCEInfoApplicationDir: /exper-sw/dteamsoft
GlueCEInfoDataDir: unset
GlueChunkKey: GlueCEUniqueID=ce02.lip.pt:2119/jobmanager-lcgsge-dteamgrid
GlueSchemaVersionMajor: 1
GlueSchemaVersionMinor: 3
(...)
dn:
GlueVOViewLocalID=dteam,GlueCEUniqueID=ce02.lip.pt:2119/jobmanager-lcgsge-
dteamgrid_x86_64,Mds-Vo-name=resource,o=grid
objectClass: GlueCETop
objectClass: GlueVOView
objectClass: GlueCEInfo
objectClass: GlueCEState
objectClass: GlueCEAccessControlBase
objectClass: GlueCEPolicy
objectClass: GlueKey
objectClass: GlueSchemaVersion
GlueVOViewLocalID: dteam
GlueCEAccessControlBaseRule: VO:dteam
GlueCEStateRunningJobs: 0
GlueCEStateWaitingJobs: 0
GlueCEStateTotalJobs: 0
GlueCEStateFreeJobSlots: 215
GlueCEStateEstimatedResponseTime: 0
GlueCEStateWorstResponseTime: 0
GlueCEInfoDefaultSE: se05.lip.pt
GlueCEInfoApplicationDir: /exper-sw/dteamsoft
GlueCEInfoDataDir: unset
GlueChunkKey:
GlueCEUniqueID=ce02.lip.pt:2119/jobmanager-lcgsge-dteamgrid_x86_
64
GlueSchemaVersionMajor: 1
GlueSchemaVersionMinor: 3
### Question 2 ###
Is this the intended behaviour? If not, is there a way to proper
configure it?
Thanks in Advance
Cheers
Goncalo
|