Re: [Condor-users] Condorg: Permission denied (publickey,password)

[Vineet Chadha <vineet.chadha@xxxxxxxxx>]

Erik,

Thanks for quick reply.  you were right that condor_submit I was
trying through  jobmanager-pbs and globus-job-run thrugh jobmanger.
Apparently, my PBS setup is not correct ...I am checking on same.
I have another problem to deal with.  I am able to submit condor job
through jobmanager  ( let say following :

executable = /bin/date
globusscheduler = xxx.xxx.xxx.xxx/jobmanager
universe = globus
error = globus-test.err
output = globus-test.out
log = globus-test.log
queue = submit

Above job description works fine.

In my server where i am submitting a job, I have created following
script called "test"
/bin/echo hello
test is executable.

Now following doesnot work

executable = /home/foo/bin/test
globusscheduler = xxx.xxx.xxx.xxx/jobmanager
universe = globus
error = globus-test.err
output = globus-test.out
log = globus-test.log
queue = submit

where as
globus-job-run xxx.xxx.xxx.edu/jobmanager /home/vam000/bin/test
works fine

Here is  output of condor_q -l of mentioned job


MyType = "Job"
TargetType = "Machine"
ClusterId = 137
QDate = 1117744545
CompletionDate = 0
Owner = "vam002"
RemoteWallClockTime = 0.000000
LocalUserCpu = 0.000000
LocalSysCpu = 0.000000
RemoteUserCpu = 0.000000
RemoteSysCpu = 0.000000
ExitStatus = 0
NumCkpts = 0
NumRestarts = 0
CommittedTime = 0
TotalSuspensions = 0
LastSuspensionTime = 0
CumulativeSuspensionTime = 0
ExitBySignal = FALSE
CondorVersion = "$CondorVersion: 6.6.9 Mar 10 2005 $"
CondorPlatform = "$CondorPlatform: I386-LINUX_RH9 $"
RootDir = "/"
Iwd = "/home/vam002"
JobUniverse = 9
Cmd = "/home/vam000/bin/test"
MinHosts = 1
MaxHosts = 1
CurrentHosts = 0
WantRemoteSyscalls = FALSE
WantCheckpoint = FALSE
x509userproxysubject = "/O=ufl.edu/OU=UF_Grid/OU=xxx.xxx.edu/CN=vam002/CN=proxy"
x509userproxy = "/tmp/x509up_u511"
JobPrio = 0
User = "xxxx@xxxxxxxxxxxxxxx"
NiceUser = FALSE
Env = ""
JobNotification = 2
UserLog = "/home/vam002/meme-test.log"
CoreSize = 0
KillSig = "SIGTERM"
Rank = 0.000000
In = "/dev/null"
TransferIn = FALSE
Out = "meme-test.out"
Err = "meme-test.err"
BufferSize = 524288
BufferBlockSize = 32768
ShouldTransferFiles = "NO"
TransferFiles = "NEVER"
ImageSize = 0
ExecutableSize = 0
DiskUsage = 0
Requirements = TRUE
FileSystemDomain = "xxx.xxx.xxx.xxx"
PeriodicHold = FALSE
PeriodicRelease = FALSE
PeriodicRemove = FALSE
OnExitHold = FALSE
OnExitRemove = TRUE
LeaveJobInQueue = FALSE
Args = ""
GlobusResource = "xxx.xxx.xxxx/jobmanager"
WantClaiming = FALSE
GlobusResubmit = FALSE
ProcId = 0
NumGlobusSubmits = 1
GlobusStatus = 64
GlobusContactString = "X"
GlobusGramVersion = 0
JobStatus = 5
HoldReason = "Globus error 43: the job manager failed to stage the executable"
HoldReasonCode = 2
HoldReasonSubCode = 43
ReleaseReason = UNDEFINED
EnteredCurrentStatus = 1117744559
NumSystemHolds = 1
Managed = FALSE
ServerTime = 1117745252

Please help

Regards,
Vineet

On 6/2/05, Erik Paulson <epaulson@xxxxxxxxxxx> wrote:
> On Thu, Jun 02, 2005 at 01:29:33PM -0400, Vineet Chadha wrote:
> > Hi all ,
> >
> > When i try to submit a globus job through condor-g  I get following
> > message in error file:
> >
> > Permission denied, please try again.
> > Permission denied, please try again.
> > Permission denied (publickey,password).
> >
> 
> Your PBS setup is wrong - those error messages are from ssh, and it's
> from PBS trying to ssh to a compute node to run your job. You need to
> setup password-less logins for it to work.
> 
> > To give the preivew of my job desciption file :
> >
> > executable = /bin/ls
> > globusscheduler = XXX/jobmanager-pbs
> > universe = globus
> > grid_type = gt3
> > error = globus-test.$(CLUSTER).err
> > output = globus-test.$(CLUSTER).out
> > should_transfer_files = IF_NEEDED
> > when_to_transfer_output = ON_EXIT
> > log = globus-test.log
> > queue
> >
> > And grid manager log for that user is as  follows
> >
> <...>
> >
> >
> > I am able to submit jobs through globus-job-run but its giving
> > permission denied message  with condor-g. Please help.
> >
> 
> globus-job-run to jobmanager-pbs, or to jobmanager-fork? And are
> you submitting with the same GSI subject name? I'm not sure
> Condor-G would do different than globus-job-run (maybe some file
> staging stuff or something causes the jobmanager to take a different
> codepath)
> 
> If it is to jobmanager-pbs and the same GSI user, would you submit the
> job again and give us the output of 'condor_q -l', as well as the
> exact arguments you use for globus-job-run?
> 
> -Erik
> 


-- 
VIneet chadha