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

Re: [Condor-users] Performance Counter Error



There is no way to avoid the error, because it shouldn't be happening in the first  place.  

The code is calling ReqQueryValueEx(HKEY_PERFORMANCE_DATA, PERF_PROCESS_LIST ),  the function is returning success, but the data appears to be partly invalid - it has the correct starting bytes, but there is a 0 in the field that shows the size of one of the counters. (the first one that we check)

The message that you  are seeing is there just so that if some future version of Windows changes the structure of the performance data in a way that violates some assumption in the code, we get a message rather than a random crash.    The code is checking to see if sizeof(perf_counter) == 8 on the current platform.  If this fails, I would expect it to fail all of the time on a given version of Windows.

Please let us know if you see this happen again, but I think it's very unlikely that its a bug in the Condor code and very very unlikely that it's a configuration problem.   It might be a bad DIMM, or a Windows bug, but I still think that you should check for a rootkit type virus - a virus attempting to edit itself out of the process list might cause this symptom.

-tj




On 7/12/2011 11:30 AM, kschwarz@xxxxxxxxxxxxxx wrote:

TJ,

The second explanation (virus) is an unlikely possibility because our company installs an antivirus software in all desktops but could happen if the virus definition does not have its definition yet.

Is there any way to avoid this error? We do not have many Windows 7 machines yet, but it will be a problem when more win7 desktops will be deployed in the company and an end user, that does not have rights to start the Condor service, needs to run jobs.

Thanks,
Klaus




"John (TJ) Knoeller" <johnkn@xxxxxxxxxxx>
Sent by: condor-users-bounces@xxxxxxxxxxx

12/07/2011 12:33

Please respond to
Condor-Users Mail List <condor-users@xxxxxxxxxxx>

To
Condor-Users Mail List <condor-users@xxxxxxxxxxx>
cc

Subject
Re: [Condor-users] Performance Counter Error







There's nothing much you can do about this unfortunately.   This is a truly bizarre thing to be intermittent.
The message is a bit misleading, you are not expected to have a key at that point in the registry.  Its just
that if you DO have a key there, it's value should be 0.

There's a Windows API  that we query that returns a structure and another API that returns some information
describing the format of that structure.  That second API is returning unexpected values for the formatting
information.

The thing is, that formatting information isn't allowed to change at runtime.  In practice it changes between
versions of Windows, but doesn't change within a version.   So the error you are seeing simply CANT be
intermittent.  

I can think of only 2 possible explanations.

1)  Something in Condor_master is randomly overwriting memory and it happened to hit this formatting
information - unlikely.

2) You have a virus that is trying to hide itself and the code isn't quite right.

-tj






On 7/12/2011 10:05 AM,
kschwarz@xxxxxxxxxxxxxx wrote:

HI,


When Condor started this morning on my Win 7 Desktop, it aborted and logged the following error message in the MasterLog file:


07/12/11 07:01:50 DaemonCore: command socket at <10.3.29.182:55868>

07/12/11 07:01:50 DaemonCore: private command socket at <10.3.29.182:55868>

07/12/11 07:01:50 Setting maximum accepts per cycle 4.

07/12/11 07:01:51 Started DaemonCore process "C:\condor/bin/condor_schedd.exe", pid and pgroup = 3008

07/12/11 07:01:52 Started DaemonCore process "C:\condor/bin/condor_startd.exe", pid and pgroup = 2268

07/12/11 07:01:55 Started DaemonCore process "C:\condor/bin/condor_kbdd.exe", pid and pgroup = 3152

07/12/11 07:01:55 ERROR "Unexpected performance counter size for total CPU: 0 (expected 8); Registry key HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\PerfProc\Performance must have 'Disable Performance Counters' value of 0 or no such value" at line 2648 in file c:\condor\execute\dir_4052\userdir\src\condor_procapi\procapi.cpp

07/12/11 07:01:55 Sent SIGKILL to KBDD (pid 3152) and all its children.

07/12/11 07:01:55 Sent SIGKILL to SCHEDD (pid 3008) and all its children.

07/12/11 07:01:55 Sent SIGKILL to STARTD (pid 2268) and all its children.


There isn´t an entry
Disable Performance Counters in this registry path.

I am using version


C:\Users\kschwarz>condor_version

$CondorVersion: 7.6.0 Apr 16 2011 BuildID: 327460 $

$CondorPlatform: x86_winnt_5.1 $


C:\Users\kschwarz>


Trying to start it again, the error disappeared and Condor is running well.


What should I do to prevent this to happening again?


Thanks,

Klaus


This message is intended solely for the use of its addressee and may contain privileged or confidential information. All information contained herein shall be treated as confidential and shall not be disclosed to any third party without Embraer’s prior written approval. If you are not the addressee you should not distribute, copy or file this message. In this case, please notify the sender and destroy its contents immediately.
Esta mensagem é para uso exclusivo de seu destinatário e pode conter informações privilegiadas e confidenciais. Todas as informações aqui contidas devem ser tratadas como confidenciais e não devem ser divulgadas a terceiros sem o prévio consentimento por escrito da Embraer. Se você não é o destinatário não deve distribuir, copiar ou arquivar a mensagem. Neste caso, por favor, notifique o remetente da mesma e destrua imediatamente a mensagem.



_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to
condor-users-request@xxxxxxxxxxx with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/condor-users/
_______________________________________________
Condor-users mailing list
To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/condor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/condor-users/



This message is intended solely for the use of its addressee and may contain privileged or confidential information. All information contained herein shall be treated as confidential and shall not be disclosed to any third party without Embraer’s prior written approval. If you are not the addressee you should not distribute, copy or file this message. In this case, please notify the sender and destroy its contents immediately.
Esta mensagem é para uso exclusivo de seu destinatário e pode conter informações privilegiadas e confidenciais. Todas as informações aqui contidas devem ser tratadas como confidenciais e não devem ser divulgadas a terceiros sem o prévio consentimento por escrito da Embraer. Se você não é o destinatário não deve distribuir, copiar ou arquivar a mensagem. Neste caso, por favor, notifique o remetente da mesma e destrua imediatamente a mensagem.
_______________________________________________ Condor-users mailing list To unsubscribe, send a message to condor-users-request@xxxxxxxxxxx with a subject: Unsubscribe You can also unsubscribe by visiting https://lists.cs.wisc.edu/mailman/listinfo/condor-users The archives can be found at: https://lists.cs.wisc.edu/archive/condor-users/