[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] Jobs are Executed Only on the Central Manager



On Mon, Sep 18, 2006 at 01:52:13PM +0800, leo@xxxxxxxxxxxxxxxxxxxxx wrote:
> 
<...>
> After all the jobs are executed:
> 
<...>

It'd be more helpful to see the condor_status while you've got jobs
submitted - but in all likelyhood the problem is the same.

The only machine that is available to run jobs is your central manager,
because Condor has detected keyboard or console activity on all of the
other machines, and they're in "Owner" state and not willing to run 
jobs. Are people sitting at the machine, or are they logged in via ssh?

Change your START expression to not reference the keyboard, or get
people to leave the machines idle and your jobs will run on all the machines.

-Erik

> 
> ###########################################3
> [condor@phys-ugradlab01 test]$ condor_status -l
> MyType = "Machine"
> TargetType = "Job"
> Name = "nucleus.cluster.physics.msuiit.edu.ph"
> Machine = "nucleus.cluster.physics.msuiit.edu.ph"
> KeyboardIdle = 0
> ConsoleIdle = 0
> State = "Owner"
> Activity = "Idle"
> Start = ((KeyboardIdle > 15 * 60) && (((LoadAvg - CondorLoadAvg) <=
> 0.300000) || (State != "Unclaimed" && State != "Owner")))
> Requirements = (START) && (IsValidCheckpointPlatform)
> 
> MyType = "Machine"
> TargetType = "Job"
> Name = "phys-ugradlab01.physics.msuiit.edu.ph"
> Machine = "phys-ugradlab01.physics.msuiit.edu.ph"
> KeyboardIdle = 600
> ConsoleIdle = 600
> State = "Unclaimed"
> Activity = "Idle"
> Start = TRUE
> Requirements = (START) && (IsValidCheckpointPlatform)
> 
> MyType = "Machine"
> TargetType = "Job"
> Name = "phys-ugradlab02.physics.msuiit.edu.ph"
> Machine = "phys-ugradlab02.physics.msuiit.edu.ph"
> KeyboardIdle = 0
> ConsoleIdle = 0
> State = "Owner"
> Activity = "Idle"
> Start = ((KeyboardIdle > 15 * 60) && (((LoadAvg - CondorLoadAvg) <=
> 0.300000) || (State != "Unclaimed" && State != "Owner")))
> Requirements = (START) && (IsValidCheckpointPlatform)
> 
> MyType = "Machine"
> TargetType = "Job"
> Name = "phys-ugradlab03.physics.msuiit.edu.ph"
> Machine = "phys-ugradlab03.physics.msuiit.edu.ph"
> KeyboardIdle = 0
> ConsoleIdle = 0
> State = "Owner"
> Activity = "Idle"
> Start = ((KeyboardIdle > 15 * 60) && (((LoadAvg - CondorLoadAvg) <=
> 0.300000) || (State != "Unclaimed" && State != "Owner")))
> Requirements = (START) && (IsValidCheckpointPlatform)