Message boards : Number crunching : Error while computing/too many errors
Message board moderation

To post messages, you must log in.

AuthorMessage
Jim Martin

Send message
Joined: 26 Jul 05
Posts: 17
Credit: 1,142,450
RAC: 2
Message 32987 - Posted: 4 Nov 2017, 21:57:09 UTC
Last modified: 4 Nov 2017, 22:02:43 UTC

I am getting many "Error while computing" errors (Too many total results).
The pass rate is less than 10%. My system is a Dell Latitude #7240 (4cpu's),
with Windows 7, BOINC ver. 7.8.3, and VBox 5.1.3r.

The errors occur, even when all other WU's have been "suspended".

Any ideas?
ID: 32987 · Report as offensive     Reply Quote
ivan
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar

Send message
Joined: 29 Aug 05
Posts: 997
Credit: 6,264,307
RAC: 71
Message 32989 - Posted: 4 Nov 2017, 22:29:36 UTC - in response to Message 32987.  

VirtualBox 5.1.30 is not the latest version; try updating (https://www.virtualbox.org/wiki/Downloads). Make sure you download and install the VirtualBox etensions
(http://download.virtualbox.org/virtualbox/5.2.0/Oracle_VM_VirtualBox_Extension_Pack-5.2.0-118431.vbox-extpack), and verify that virtual machines are enabled in your BIOS.
ID: 32989 · Report as offensive     Reply Quote
Jim Martin

Send message
Joined: 26 Jul 05
Posts: 17
Credit: 1,142,450
RAC: 2
Message 32990 - Posted: 4 Nov 2017, 23:14:27 UTC - in response to Message 32989.  

Ivan -- Thanks, for the advice. I'll update.
ID: 32990 · Report as offensive     Reply Quote
Jim Martin

Send message
Joined: 26 Jul 05
Posts: 17
Credit: 1,142,450
RAC: 2
Message 33016 - Posted: 8 Nov 2017, 0:12:43 UTC - in response to Message 32990.  
Last modified: 8 Nov 2017, 0:19:55 UTC

Ivan -- I've run both LHCb and CMS WU's, with VirtualBox 5.2.0. etc., and still receive the same errors, as before. Shall I keep running them, or await more info. from you?
ID: 33016 · Report as offensive     Reply Quote
ivan
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar

Send message
Joined: 29 Aug 05
Posts: 997
Credit: 6,264,307
RAC: 71
Message 33020 - Posted: 8 Nov 2017, 8:23:10 UTC - in response to Message 33016.  

Ivan -- I've run both LHCb and CMS WU's, with VirtualBox 5.2.0. etc., and still receive the same errors, as before. Shall I keep running them, or await more info. from you?

I'm afraid you will be getting compute errors with CMS jobs at the moment as the WMAgent failed last night and the job queue drained (so no jobs from condor). Don't run CMS tasks until the job activity graph springs back into life.
Sorry 'bout that...
ID: 33020 · Report as offensive     Reply Quote
Profile adrianxw

Send message
Joined: 29 Sep 04
Posts: 187
Credit: 705,487
RAC: 0
Message 33055 - Posted: 13 Nov 2017, 7:51:32 UTC

I had a small group of "Error while computing" last weekend. An LHCb, a CMS and two Theory Sim's.

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 33055 · Report as offensive     Reply Quote
Jim Martin

Send message
Joined: 26 Jul 05
Posts: 17
Credit: 1,142,450
RAC: 2
Message 33111 - Posted: 22 Nov 2017, 0:45:57 UTC

"Error while computing" errors, have resumed, with the latest WU's.
CMS and LHCb.
ID: 33111 · Report as offensive     Reply Quote
Profile Magic Quantum Mechanic
Avatar

Send message
Joined: 24 Oct 04
Posts: 1114
Credit: 49,501,728
RAC: 4,157
Message 33112 - Posted: 22 Nov 2017, 1:51:59 UTC - in response to Message 33111.  
Last modified: 22 Nov 2017, 1:54:57 UTC

"Error while computing" errors, have resumed, with the latest WU's.
CMS and LHCb.


With you CMS errors I see the first thing on the tasks stderr is
Error creating VirtualBox instance

You do have the newest version of both VB and Boinc and Oracle says that new version of VB did have a problem with Linux but didn't mention Windows 7 or 10

https://www.virtualbox.org/wiki/Downloads

There are several usual problems when starting a new version of VB

Did you do the d/l from here with both Boinc and VB or did you d/l VB separately from the Wiki link?

Over all the years I have always found upgrading to the new Boinc is best done first and then d/l the new VB from the Wiki link with the Extension Pack

THEN reboot before trying to start a new LHC-VB task

Another problem is not having a fast enough internet d/l speed to start the tasks and get to the *Credentials* and server connection at Cern and finally the HTCondor ping BEFORE the time is up to do that (10 minutes with CMS and 20 or less with the Theory tasks)

Another problem can be checked on your VB Manager
Go to the VB Manager and then File/VB Media Manager/ and in that box you may find some vdi's that need to be removed since they can mess up the new tasks trying to get a slot to use.

This is what you do not want to see.....and the good and the bad there



And you can check in that VB Manager for other things such as if that Extension Pack has that Green check mark next to it with that newest version.

I tend to watch the VM Console for the tasks as they start so I know for sure they actually got to HTCondor ping so they don't waste time and end up a Computer Error.

Once you get used to these tasks you can usually find any problems on your end......now the Cern server and tasks are another story
Volunteer Mad Scientist For Life
ID: 33112 · Report as offensive     Reply Quote
Profile adrianxw

Send message
Joined: 29 Sep 04
Posts: 187
Credit: 705,487
RAC: 0
Message 33123 - Posted: 23 Nov 2017, 10:22:04 UTC

>>> The filename or extension is too long.
>>> (0xce) - exit code 206 (0xce)

I'm seeing a few exits again now. Saw that in the error log.

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 33123 · Report as offensive     Reply Quote
ivan
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Project scientist
Avatar

Send message
Joined: 29 Aug 05
Posts: 997
Credit: 6,264,307
RAC: 71
Message 33125 - Posted: 23 Nov 2017, 13:32:51 UTC - in response to Message 33123.  

There is an ongoing intervention with some CMS systems, expected to finish in an hour or two. That's probably not causing that particular error though, I wouldn't have thought.
ID: 33125 · Report as offensive     Reply Quote
Jim Martin

Send message
Joined: 26 Jul 05
Posts: 17
Credit: 1,142,450
RAC: 2
Message 33165 - Posted: 29 Nov 2017, 23:17:19 UTC

I think the problem has been resolved. Firstly, I reboot the system, before
starting the first LHC WU; whether that has any affect on preventing
errors, is not known, but, no errors have followed. Also, as before, I don't
have any other WU's running, except the particular LHC one. Any time I need
to check personal email, etc., I "pause" the running WU.
ID: 33165 · Report as offensive     Reply Quote
marmot
Avatar

Send message
Joined: 5 Nov 15
Posts: 144
Credit: 6,301,268
RAC: 0
Message 33177 - Posted: 30 Nov 2017, 15:37:57 UTC - in response to Message 33112.  






This looked like all of my servers VBox media manager lists and those had to be deleted 1 at a time. No multi-selection. Upgrading to the latest VBox didn't help (but that machine seems to save LHC VM's more quickly on suspend now).

A solution discovered, in a fit of frustration, was to suspend all WU's (4 at a time), close BOINC and use a process manager to kill the VBoxSVC.exe service. Once the VBox Manager interface is reopened, the environment is cleaned up. All the broken links to the BOINC data\slot *.vdi files are gone without having to delete them all and without having to reboot the machine.

The numbers of broken links to slot vdi's (30, 50, 60 per day) appear to be reduced to a few per day by using Process Hacker and forcing VBoxManager.exe to normal priority from the default idle priority in the assumption it was a time-out issue on a computer where all cores are running at maximum. (Seen crypto wallet's and other apps crash regularly until their priority was raised).

This could be a coincidence so I'd like to know if it helps anyone else.
ID: 33177 · Report as offensive     Reply Quote
Profile adrianxw

Send message
Joined: 29 Sep 04
Posts: 187
Credit: 705,487
RAC: 0
Message 34093 - Posted: 26 Jan 2018, 9:51:19 UTC

Picked up 3 recent Error while computing errors again.

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 34093 · Report as offensive     Reply Quote
computezrmle
Volunteer moderator
Volunteer developer
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 15 Jun 08
Posts: 2386
Credit: 222,874,623
RAC: 137,191
Message 34095 - Posted: 26 Jan 2018, 10:15:55 UTC - in response to Message 34093.  

Picked up 3 recent Error while computing errors again.

You may examine your logs.



https://lhcathome.cern.ch/lhcathome/result.php?resultid=173873928
2018-01-18 21:20:25 (15796): Guest Log: [DEBUG] nc: connect to vccondor01.cern.ch port 9618 (tcp) timed out: Operation now in progress
2018-01-18 21:20:25 (15796): Guest Log: [DEBUG] 1
2018-01-18 21:20:25 (15796): Guest Log: [ERROR] Could not connect to Condor server on port 9618
2018-01-18 21:20:25 (15796): Guest Log: [INFO] Shutting Down.

Most likely a problem with the Condor server.



https://lhcathome.cern.ch/lhcathome/result.php?resultid=174134896
Most likely: No jobs available at startup time -> VM paused -> Timeout reached after wakeup



https://lhcathome.cern.ch/lhcathome/result.php?resultid=174107552
2018-01-21 00:31:48 (8360): VM state change detected. (old = 'Running', new = 'Paused')
2018-01-21 17:04:04 (8360): VM state change detected. (old = 'Paused', new = 'Running')

This VM was paused for too long and as it did not finished at least 1 job before it was paused, it was treated as an error.
ID: 34095 · Report as offensive     Reply Quote
Profile adrianxw

Send message
Joined: 29 Sep 04
Posts: 187
Credit: 705,487
RAC: 0
Message 34105 - Posted: 26 Jan 2018, 16:59:18 UTC

>>> Most likely: No jobs available at startup time

... so it trashes an in progress work unit, I doubt that.

>>> This VM was paused for too long

... same comment really, if the machine was off or busy or whatever, it trashes a work unit, I doubt that.

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 34105 · Report as offensive     Reply Quote
AuxRx

Send message
Joined: 16 Sep 17
Posts: 100
Credit: 1,618,469
RAC: 0
Message 34106 - Posted: 26 Jan 2018, 17:47:01 UTC - in response to Message 34105.  

... same comment really, if the machine was off or busy or whatever, it trashes a work unit, I doubt that.


Sorry, but that's the way it is. It's all about Condor connections.
ID: 34106 · Report as offensive     Reply Quote
Profile adrianxw

Send message
Joined: 29 Sep 04
Posts: 187
Credit: 705,487
RAC: 0
Message 34109 - Posted: 26 Jan 2018, 19:43:20 UTC

If the server cannot support its own running tasks on remote worker units that represents a serious problem. Not just support, but fail them?

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 34109 · Report as offensive     Reply Quote

Message boards : Number crunching : Error while computing/too many errors


©2024 CERN