Development for LHC@home

We are currently testing multi-core jobs for CMS@Home. Note that these will only run in -dev as the main project does not currently allow you to select multi-core VMs. We currently have 2-core and 4-core tasks in the queue, so please try selecting 4-core in your machine preferences, and let us know how it works.

---

duplicate, sorry.

Yes, 4-Core works well since a few days.
https://lhcathomedev.cern.ch/lhcathome-dev/results.php?hostid=5116

What is the process cmsExternalGene doing?
Now I see 4 of those processses on my 4-core VM using up to 100% cpu each.
Do they the event processing and if yes: Does it mean I'm processing 4 jobs concurrently.
That would not make much sense.

dupe

tripled

Still running after 13.5 hours the first 4 jobs:



On slow but a bit faster than this machine, it could happen, that the second 4 jobs are been killed by the 18 hours deadline.
Why not run 1 single job 4 times faster like you do with the dual core tasks. 1 cmsRun using 200% cpu and twice as fast.
https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3311582

Still running after 13.5 hours the first 4 jobs:
...
...
On slow but a bit faster than this machine, it could happen, that the second 4 jobs are been killed by the 18 hours deadline.
Why not run 1 single job 4 times faster like you do with the dual core tasks. 1 cmsRun using 200% cpu and twice as fast.
https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3311582

Just in time:

Run time 17 hours 32 min 57 sec
Only 4 concurrently running jobs.

Crystal,
seeing the same: 4 cmsExternalGene... 153 minutes hours on 4-Cores!

Crystal,
seeing the same: 4 cmsExternalGene... 153 minutes hours on 4-Cores!

Your task https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3311748 seems to be killed by the 18 hours deadline.

Your result ends with:
2024-03-26 16:45:34 (33976): Status Report: Elapsed Time: '60000.000000'
2024-03-26 16:45:34 (33976): Status Report: CPU Time: '216839.343750'
2024-03-26 18:05:37 (33976): Powering off VM.
2024-03-26 18:05:38 (33976): Successfully stopped VM.
2024-03-26 18:05:38 (33976): Deregistering VM. (boinc_9f8cb45e4f80768f, slot#27)
2024-03-26 18:05:38 (33976): Removing network bandwidth throttle group from VM.
2024-03-26 18:05:38 (33976): Removing VM from VirtualBox.

Mine with:
2024-03-26 13:06:15 (1020): Guest Log: [INFO] glidein exited with return value 0.
2024-03-26 13:06:15 (1020): Guest Log: [INFO] Shutting Down.
2024-03-26 13:06:15 (1020): VM Completion File Detected.
2024-03-26 13:06:15 (1020): VM Completion Message: glidein exited with return value 0.
.
2024-03-26 13:06:15 (1020): Powering off VM.
2024-03-26 13:06:16 (1020): Successfully stopped VM.
2024-03-26 13:06:16 (1020): Deregistering VM. (boinc_15fe7a25adba3060, slot#0)
2024-03-26 13:06:16 (1020): Removing network bandwidth throttle group from VM.
2024-03-26 13:06:16 (1020): Removing VM from VirtualBox.

This was luck ;-))
Computer ID 4639
Laufzeit 18 Stunden 0 min. 54 sek.
CPU Zeit 2 Tage 17 Stunden 22 min. 43 sek.
Prüfungsstatus Gültig
Punkte 3,721.09

This was luck ;-))

No luck for CMS, cause the running cms jobs did not return a result.
Your task was valid BOINC-wise, but a valid cms-job should end with: VM Completion Message: glidein exited with return value 0.
0 means no error. Your job ended by a shutdown given by vboxwrapper after 18 hours wall clock time.

We are currently testing multi-core jobs for CMS@Home. Note that these will only run in -dev as the main project does not currently allow you to select multi-core VMs. We currently have 2-core and 4-core tasks in the queue, so please try selecting 4-core in your machine preferences, and let us know how it works.

It has been almost a full month since these tasks were released and I have been running two computers (an i7-4790 CPU and an i7-8650U CPU) with nothing except a single 4-thread task on each of them the whole time, and one other computer (a Xeon E5-2699 v3 CPU) with one 4-thread LHC task running along with a bunch of other work from different BOINC projects, and my task history says that there have only been a couple of failures that whole time. And I'm pretty sure those failures were my own fault from forcing the computer to reboot without letting BOINC stop the task properly first.