Advanced search

Message boards : Number crunching : All acemd3 tasks are defective

Author Message
Magiceye04
Send message
Joined: 1 Apr 09
Posts: 24
Credit: 67,905,687
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwat
Message 59637 - Posted: 21 Dec 2022 | 19:26:08 UTC

Not only for me, all users before also got these defective WUs.
Could please someone stop sending out this waste?

Pop Piasa
Avatar
Send message
Joined: 8 Aug 19
Posts: 252
Credit: 458,054,251
RAC: 0
Level
Gln
Scientific publications
watwat
Message 59648 - Posted: 24 Dec 2022 | 1:41:24 UTC - in response to Message 59637.
Last modified: 24 Dec 2022 | 2:30:24 UTC

Could please someone stop sending out this waste?


May I humbly submit...
If Adria learns from this setback it is not a waste, it becomes a step that had to be taken to move on. This is how research is conducted. Trial and error. It is not being presented for our entertainment as competitors, so it carries the caveat that experiments sometimes fail.

(I hope that wasn't put too bluntly, I think I maybe was channeling Karl Popper there.) 😉
____________
"Together we crunch
To check out a hunch
And wish all our credit
Could just buy us lunch"


Piasa Tribe - Illini Nation

Magiceye04
Send message
Joined: 1 Apr 09
Posts: 24
Credit: 67,905,687
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwat
Message 59661 - Posted: 27 Dec 2022 | 12:00:51 UTC

Of course an experiment can go wrong. But as I requested - then the experiment shall be stopped. In best case the experiment is pre-tested as beta-version with just a few WUs before it starts for all users.

Keith Myers
Send message
Joined: 13 Dec 17
Posts: 1354
Credit: 7,814,235,809
RAC: 8,963,340
Level
Tyr
Scientific publications
watwatwatwatwat
Message 60888 - Posted: 14 Dec 2023 | 22:35:23 UTC

How many other Linux users are having difficulty running this sparse release of acemd3 tasks?

I haven't been successful on any of my Linux hosts.

Looking at the other wingmen that are offered the tasks, I see similar failures from other Linux wingmen.

And usually the task is finally completed by a Windows host. They seem to have no issues with the tasks.

I've never had any issues with the acemd3 task in the past.

And I have no issues with the ATMbeta tasks when they are offered.

Profile ServicEnginIC
Avatar
Send message
Joined: 24 Sep 10
Posts: 581
Credit: 10,066,573,373
RAC: 16,148,289
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 60889 - Posted: 15 Dec 2023 | 5:46:21 UTC - in response to Message 60888.

Exactly the same here on my Linux hosts for current ACEMD 3 tasks.
They all fail after a few seconds, with no clue about the reason.

Stderr output

<core_client_version>7.20.5</core_client_version>
<![CDATA[
<message>
process exited with code 195 (0xc3, -61)</message>
<stderr_txt>
04:36:04 (51194): wrapper (7.7.26016): starting
04:36:21 (51194): wrapper (7.7.26016): starting
04:36:21 (51194): wrapper: running bin/acemd3 (--boinc --device 0)
04:36:22 (51194): bin/acemd3 exited; CPU time 0.000000
04:36:22 (51194): app exit status: 0x1
04:36:22 (51194): called boinc_finish(195)

And eventually, the same WU succeeds when a new task arrives a Windows host.
I've reset GPUGRID project at BOINC Manager, and no variation.
No problem witn ATMbeta tasks on the same hosts.

Bedrich Hajek
Send message
Joined: 28 Mar 09
Posts: 485
Credit: 11,282,638,355
RAC: 10,864,042
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 60890 - Posted: 15 Dec 2023 | 6:35:24 UTC - in response to Message 60889.

Exactly the same here on my Linux hosts for current ACEMD 3 tasks.
They all fail after a few seconds, with no clue about the reason.

Stderr output

<core_client_version>7.20.5</core_client_version>
<![CDATA[
<message>
process exited with code 195 (0xc3, -61)</message>
<stderr_txt>
04:36:04 (51194): wrapper (7.7.26016): starting
04:36:21 (51194): wrapper (7.7.26016): starting
04:36:21 (51194): wrapper: running bin/acemd3 (--boinc --device 0)
04:36:22 (51194): bin/acemd3 exited; CPU time 0.000000
04:36:22 (51194): app exit status: 0x1
04:36:22 (51194): called boinc_finish(195)



And eventually, the same WU succeeds when a new task arrives a Windows host.
I've reset GPUGRID project at BOINC Manager, and no variation.
No problem witn ATMbeta tasks on the same hosts.


I have seen this before, a few years ago, on few occasions. It happened when the license for the Linux app expired, and the project management forgot to renew it. Sometimes, it was the other way, the Windows license expired, and all the Windows computers would fail and the Linux computers would succeed.

If you move the clock on your computer to before the expiration date of the app, before the task starts, the task would succeed, but boinc interface would be "unresponsive"......

Ian&Steve C.
Avatar
Send message
Joined: 21 Feb 20
Posts: 1076
Credit: 40,231,533,983
RAC: 98
Level
Trp
Scientific publications
wat
Message 60891 - Posted: 15 Dec 2023 | 13:03:34 UTC - in response to Message 60890.

you're right

$ ./acemd3
#
# ACEMD version 3.5.1
#
# Copyright (C) 2017-2022 Acellera (www.acellera.com)
#
# When publishing, please cite:
# ACEMD: Accelerating Biomolecular Dynamics in the Microsecond Time Scale
# M. J. Harvey, G. Giupponi and G. De Fabritiis,
# J Chem. Theory. Comput. 2009 5(6), pp1632-1639
# DOI: 10.1021/ct9000685
#
# Arguments:
# input: input
# platform:
# device:
# ncpus:
# precision: mixed
#
# Licence:
# Check floating licence:
# ACELLERA_LICENCE_SERVER -- not defined
# ACELLERA_LICENSE_SERVER -- not defined
# Check node-locked licence:
# ACELLERA_LICENCE_FILE -- not defined
# ACELLERA_LICENSE_FILE -- not defined
# /opt/acellera/licence.dat -- DENIED (Unable to locate target file)
# /opt/acellera/license.dat -- DENIED (Unable to locate target file)
# /home/ian/.acellera/licence.dat -- DENIED (Unable to locate target file)
# /home/ian/.acellera/license.dat -- DENIED (Unable to locate target file)
#
# This version of ACEMD has expired!
# You should update to the latest version.
# If you want to continue using this version, you need a licence.


____________

Keith Myers
Send message
Joined: 13 Dec 17
Posts: 1354
Credit: 7,814,235,809
RAC: 8,963,340
Level
Tyr
Scientific publications
watwatwatwatwat
Message 60892 - Posted: 15 Dec 2023 | 20:01:03 UTC

Gianni replied that they are on it for the fix.

Erich56
Send message
Joined: 1 Jan 15
Posts: 1132
Credit: 10,643,355,840
RAC: 24,643,273
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwat
Message 60893 - Posted: 16 Dec 2023 | 8:12:52 UTC - in response to Message 60890.

I have seen this before, a few years ago, on few occasions. It happened when the license for the Linux app expired, and the project management forgot to renew it. Sometimes, it was the other way, the Windows license expired, and all the Windows computers would fail and the Linux computers would succeed.

this has happened several times in the past. The team found out about the expiration of a license only after tasks started failing.
I never understood what's the problem with making a note in a calendar some time before the expiration date of a license, so that the license could be renewed in time.

Post to thread

Message boards : Number crunching : All acemd3 tasks are defective

//