[Condor-users] Windows AD domain security issue?

[Rob Stevenson <R.Stevenson@xxxxxxxxxxxxxxxxx>]

Hey All,

I've got two separate Active Directory domains with condor servers running on them.

 

When I submit from domain A, jobs are not running on domain B. I've already made sure the UID_DOMAIN is different so the jobs trying to run in domain B are trying to launch with the condor user ('condor-reuse-slot1').

 

The error below (from StarterLog.slot1) shows the errors - relevant ones being these, I think:

10/29/11 09:51:43 LogonUser(condor-reuse-slot1, ... ) failed with status 1385
10/29/11 09:51:43 ERROR "Failed to create a user nobody" at line 482 in file c:\condor\execute\dir_4228\userdir\src\condor_utils\uids.cpp

I've found that if I add condor-reuse-slot1 to the domain-admins group, my jobs run, so I'm pretty sure I'm one AD permissions issue away from success! Does anyone know the specific permissions I need to add to the condor users to enable them to run jobs without having to keep them in the domain admins group? And how do I add this (if it's not obvious!). I've seen a few posts asking similar things, but not found any specific answer - a few hacks which have worked in some situations, but nothing that felt 'right'.

 

Any ideas appreciated

 

Many thanks!

Rob

 

 

 

 

10/29/11 09:51:43 ******************************************************
10/29/11 09:51:43 ** condor_starter (CONDOR_STARTER) STARTING UP
10/29/11 09:51:43 ** C:\Condor\bin\condor_starter.exe
10/29/11 09:51:43 ** SubsystemInfo: name=STARTER type=STARTER(8) class=DAEMON(1)
10/29/11 09:51:43 ** Configuration: subsystem:STARTER local:<NONE> class:DAEMON
10/29/11 09:51:43 ** $CondorVersion: 7.6.3 Aug 17 2011 BuildID: 361356 $
10/29/11 09:51:43 ** $CondorPlatform: x86_winnt_5.1 $
10/29/11 09:51:43 ** PID = 4724
10/29/11 09:51:43 ** Log last touched 10/29 08:51:42
10/29/11 09:51:43 ******************************************************
10/29/11 09:51:43 Using config source: C:\Condor\condor_config
10/29/11 09:51:43 Using local config sources:
10/29/11 09:51:43    C:\Condor/condor_config.local
10/29/11 09:51:43 DaemonCore: command socket at <192.168.206.9:2965>
10/29/11 09:51:43 DaemonCore: private command socket at <192.168.206.9:2965>
10/29/11 09:51:43 Setting maximum accepts per cycle 4.
10/29/11 09:51:43 GLEXEC_JOB not supported on this platform; ignoring
10/29/11 09:51:43 Setting resource limits not implemented!
10/29/11 09:51:43 Communicating with shadow <192.9.201.133:4216>
10/29/11 09:51:43 Submitting machine is "pebble.hrw-uk.local"
10/29/11 09:51:43 setting the orig job name in starter
10/29/11 09:51:43 setting the orig job iwd in starter
10/29/11 09:51:43 LogonUser(condor-reuse-slot1, ... ) failed with status 1385
10/29/11 09:51:43 ERROR "Failed to create a user nobody" at line 482 in file c:\condor\execute\dir_4228\userdir\src\condor_utils\uids.cpp
10/29/11 09:51:43 ShutdownFast all jobs.
10/29/11 09:51:43 condor_read() failed: recv() returned -1, errno = 10054 , reading 5 bytes from <192.9.201.133:4222>.
10/29/11 09:51:43 IO: Failed to read packet header

---

HR Wallingford uses faxes and emails for confidential and legally privileged business communications. They do not of themselves create legal commitments. Disclosure to parties other than addressees requires our specific consent. We are not liable for unauthorised disclosures nor reliance upon them.
If you have received this message in error please advise us immediately and destroy all copies of it.

HR Wallingford Limited
Howbery Park, Wallingford, Oxfordshire, OX10 8BA, United Kingdom
Registered in England No. 02562099

---