LHC@home

Also these lines look rather odd:

2018-08-13 13:29:29 (912): Guest Log: CCooppyyiinngg  iinnppuut tf iflielse si nitnot oR uRnuAntAltalsa.s.
2018-08-13 13:29:29 (912): Guest Log: CCooppyyiinngg  iinnppuut tf iflielse si nitnot oR uRnuAntAltalsa.s.
2018-08-13 13:29:29 (912): Guest Log: Copied input files into RunAtlas.
2018-08-13 13:29:40 (912): Guest Log: Copied input files into RunAtlas.
2018-08-13 13:30:53 (912): Guest Log: copcopied the webapp to /var/www
2018-08-13 13:30:53 (912): Guest Log: ied the webapp to /var/www
2018-08-13 13:31:04 (912): Guest Log: TThhisi svm v md odeose sn otn onte ende etdo  steot uspe thutptp  hptrtopx yp
2018-08-13 13:31:04 (912): Guest Log: xy
2018-08-13 13:31:04 (912): Guest Log: AATHTEHNEAN_AP_RPORCO_CN_UNMUBMEBRE=R8
2018-08-13 13:31:04 (912): Guest Log: 8
2018-08-13 13:31:04 (912): Guest Log: SStarttianrgt iAnTgL AAS TjLoAbS.  j(oPba.nd a(IPDa=n4d0a2I4D64=64804294 t6a4s6k8I4D9=1 4t8a6s7k2I7D3=)1

I have some theories as to why those lines are garbled but you won't like them :)

This is still a problem, that we have seen sometimes by Alpha- or Beta-Tests, but never could identify or even solve it. But shure it is not a cheating script

---

Supporting BOINC, a great concept !

OK, I am convinced, no cheating going on. Most likely it's as vseven said, the "suspend if CPU usage is over xx %" was too low.

Also these lines look rather odd:

2018-08-13 13:29:29 (912): Guest Log: CCooppyyiinngg  iinnppuut tf iflielse si nitnot oR uRnuAntAltalsa.s.
2018-08-13 13:29:29 (912): Guest Log: CCooppyyiinngg  iinnppuut tf iflielse si nitnot oR uRnuAntAltalsa.s.
2018-08-13 13:29:29 (912): Guest Log: Copied input files into RunAtlas.
2018-08-13 13:29:40 (912): Guest Log: Copied input files into RunAtlas.
2018-08-13 13:30:53 (912): Guest Log: copcopied the webapp to /var/www
2018-08-13 13:30:53 (912): Guest Log: ied the webapp to /var/www
2018-08-13 13:31:04 (912): Guest Log: TThhisi svm v md odeose sn otn onte ende etdo  steot uspe thutptp  hptrtopx yp
2018-08-13 13:31:04 (912): Guest Log: xy
2018-08-13 13:31:04 (912): Guest Log: AATHTEHNEAN_AP_RPORCO_CN_UNMUBMEBRE=R8
2018-08-13 13:31:04 (912): Guest Log: 8
2018-08-13 13:31:04 (912): Guest Log: SStarttianrgt iAnTgL AAS TjLoAbS.  j(oPba.nd a(IPDa=n4d0a2I4D64=64804294 t6a4s6k8I4D9=1 4t8a6s7k2I7D3=)1

I have some theories as to why those lines are garbled but you won't like them :)

This is still a problem, that we have seen sometimes by Alpha- or Beta-Tests, but never could identify or even solve it. But shure it is not a cheating script

Nothing to worry about. It's a minor problem of vboxwrapper.
The wrapper tries to read frequently Guest logs from vbox.log out of the Logs folder in the slot directory.
If new log lines are found they are added to stderr.txt -> the result file.
The lines are rarely garbled and I never tried to discover under which circumstances this is happening.

10567661, 404 of 404 tasks failed, owner Anonymous, 100% failure rate

10567450, 1237 of 1237 tasks failed in past 4 days, owner George Bradshaw, team Gridcoin, 100% failure rate

10567464, 1360 of 1360 tasks failed in past 4 days, owner Anonymous, 100% failure rate

10563653, 212 of 212 tasks failed in past 4 days, owner Anonymous, 100% failure rate

This are typical shortrunner with:
VBoxManage.exe: error: VT-x is disabled in the BIOS for all CPU modes (VERR_VMX_MSR_ALL_VMX_DISABLED)
VT-X or AMD-V is not enabled, but...
other Computer finish them successful!!

but... other Computer finish them successful!!

You missed the point. It's not about bad tasks (tasks that cannot be computed successfully). It's about the project wasting resources sending tasks to misconfigured hosts that never return a successful result. Perhaps it's the reason for all these repeated "infrastructure failures", "bottlenecks" and stuck zombie tasks we keep hearing about.

Also these lines look rather odd:

2018-08-13 13:29:29 (912): Guest Log: CCooppyyiinngg  iinnppuut tf iflielse si nitnot oR uRnuAntAltalsa.s.
2018-08-13 13:29:29 (912): Guest Log: CCooppyyiinngg  iinnppuut tf iflielse si nitnot oR uRnuAntAltalsa.s.
2018-08-13 13:29:29 (912): Guest Log: Copied input files into RunAtlas.
2018-08-13 13:29:40 (912): Guest Log: Copied input files into RunAtlas.
2018-08-13 13:30:53 (912): Guest Log: copcopied the webapp to /var/www
2018-08-13 13:30:53 (912): Guest Log: ied the webapp to /var/www
2018-08-13 13:31:04 (912): Guest Log: TThhisi svm v md odeose sn otn onte ende etdo  steot uspe thutptp  hptrtopx yp
2018-08-13 13:31:04 (912): Guest Log: xy
2018-08-13 13:31:04 (912): Guest Log: AATHTEHNEAN_AP_RPORCO_CN_UNMUBMEBRE=R8
2018-08-13 13:31:04 (912): Guest Log: 8
2018-08-13 13:31:04 (912): Guest Log: SStarttianrgt iAnTgL AAS TjLoAbS.  j(oPba.nd a(IPDa=n4d0a2I4D64=64804294 t6a4s6k8I4D9=1 4t8a6s7k2I7D3=)1

I have some theories as to why those lines are garbled but you won't like them :)

This is still a problem, that we have seen sometimes by Alpha- or Beta-Tests, but never could identify or even solve it. But sure it is not a cheating script

You are correct and it was caused by a too slow internet connection with the server as I proved a few hundred times and yes that was those Alpha-Atlas tests you and I did. (I mentioned it here some time ago on one of the many threads) and you can see it doesn't happen as much here because of a couple code changes which can be seen if you watch the VM Console when they start running a new task.

(here is a copy of one from the early days before they went on to Beta)

Guest Log: CCooppyyiinngg iinnppuutt ffiilleess iinnttoo RRuunnAAttllaass..
2017-05-29 19:13:17 (6772): Guest Log: CCooppyyiinngg iinnppuutt ffiilleess iinnttoo RRuunnAAttllaass..
2017-05-29 19:13:17 (6772): Guest Log: Copied input files into RunAtlas.
2017-05-29 19:13:17 (6772): Guest Log: Copied input files into RunAtlas.
2017-05-29 19:14:27 (6772): Guest Log: ccooppiieedd tthhee wweebbaapppp ttoo //vvaarr//wwwwww
2017-05-29 19:14:27 (6772): Guest Log: ccooppiieedd tthhee wweebbaapppp ttoo //vvaarr//wwwwww
2017-05-29 19:14:27 (6772): Guest Log: TThhiiss vvmm ddooeess nnoott nneeeedd ttoo sseettuupp hhttttpp pprrooxxyy
2017-05-29 19:14:27 (6772): Guest Log: TThhiiss vvmm ddooeess nnoott nneeeedd ttoo sseettuupp hhttttpp pprrooxxyy
2017-05-29 19:14:27 (6772): Guest Log: AATTHHEENNAA__PPRROOCC__NNUUMMBBEERR==66
2017-05-29 19:14:27 (6772): Guest Log: AATTHHEENNAA__PPRROOCC__NNUUMMBBEERR==66
2017-05-29 19:14:27 (6772): Guest Log: SSttaarrttiinngg AATTLLAASS jjoobb.. ((PPaannddaaIIDD==33339988553388000055 ttaasskkIIDD==1111339977773366))
2017-05-29 19:14:27 (6772): Guest Log: SSttaarrttiinngg AATTLLAASS jjoobb.. ((PPaannddaaIIDD==33339988553388000055 ttaasskkIIDD==1111339977773366))
2017-05-29 19:19:58 (6772): Capturing screenshot.
(we save all the old test stderr's for those Alpha tests)

Just stumbled over this misconfigured host that wastes resources for more than a week.
The user is anonymous so he/she can't be contacted by normal volunteers.

Solving the problem would probably be easy as ATM it's just VT-x that has to be switched on.

If the owner doesn't take care, can the host be banned?


https://lhcathome.cern.ch/lhcathome/show_host_detail.php?hostid=10558009

VBoxManage.exe: error: VT-x is disabled in the BIOS for all CPU modes (VERR_VMX_MSR_ALL_VMX_DISABLED)

The host seems to be in the process of banning itself from VBox based tasks. It's appliication details page shows it's restricted to 1 ATLAS per day. It is still entitled to 192 LHCb and 35 Theory per day but those numbers will also decrease to 1 if it continues to fail LHCb and Theory tasks. And fail it most certainly will until VT-x is enabled.

It hasn't failed any Sixtrack lately so it's still entitled to 522 per day.

I was under the impression this "self banning" mechanism was either turned off or broken. Either I was wrong or they just fixed it or maybe turned it on recently.

After numerous Sixtrack successes it's reverted to failing Theory tasks which has decreased it's "max theory tasks allowed per day" to 7 from 35. Soon it will reduce to 1 and some some after LHCb will also reduce to 1. At that point the host is effectively a zombie that no longer devours resources, it just sleeps until it gets Sixtrack tasks which it crunches OK.

If the host doesn't get detached it will exist happily as a Sixtrack only zombie for years in which case the task downloads wasted on getting it to that state might be considered a good investment of resources.
But if it detaches and reattaches then what?... it starts wasting ATLAS downloads again?

I've stumbled on numerous hosts that fail all VBox tasks because VT-x is disabled. It would make sense to limit new hosts to 1 VBox task per day until it returns a report that indicates it meets minimum requirements such as having VT-x enabled.

It looks like this computer does nothing useful.
It only wastes resources.
https://lhcathome.cern.ch/lhcathome/show_host_detail.php?hostid=10563767

Unfortunately the project server doesn't notice it as rouge host.
I wonder if it could be banned manually until the user changes the configuration.

It looks like this computer does nothing useful.
It only wastes resources.
https://lhcathome.cern.ch/lhcathome/show_host_detail.php?hostid=10563767

Unfortunately the project server doesn't notice it as rouge host.
I wonder if it could be banned manually until the user changes the configuration.

I took a quick look and the host has more than 30 successfull jobs and only 2 or 3 bad, so something seems to be wrong with your link

---

Supporting BOINC, a great concept !

It looks like this computer does nothing useful.
It only wastes resources.
https://lhcathome.cern.ch/lhcathome/show_host_detail.php?hostid=10563767

Unfortunately the project server doesn't notice it as rouge host.
I wonder if it could be banned manually until the user changes the configuration.

I took a quick look and the host has more than 30 successfull jobs and only 2 or 3 bad, so something seems to be wrong with your link

You may look at the runtimes/CPU-times and inspect the logs.
This tells a different story.

The runtimes are far too short for ATLAS as well as for Theory.
None of the ATLAS WUs produced a HITS file.
All Theory jobs finished with an error:

Guest Log: [INFO] Job finished in slotx with 1.

You may look at the runtimes/CPU-times and inspect the logs.
This tells a different story.

The runtimes are far too short for ATLAS as well as for Theory.
None of the ATLAS WUs produced a HITS file.
All Theory jobs finished with an error:

Guest Log: [INFO] Job finished in slotx with 1.

Then we are back to an old Theme: Instead of telling "success" the Server should give back a real status. The User doesn't have a chance to recognize that his results are nothing worth

---

Supporting BOINC, a great concept !

Self-reporting my 10457508 host before anyone else does.
Lots of bother doing yesterday's Windows update with all three hosts taking several attempts to download, install, reboot, re-update but this particular host crashed on startup each time and rolled back to pre-update state. Some searching found a suggestion to reset the Bios to defaults. This got the update to work, although it still shows the Windows build as 1803 although the others have 1809 now so I'm not convinced it has actually taken the update. Running Update again shows none available and the host is useable again.

OBVIOUSLY I forgot to reallow virtualisation in Bios so all overnight tasks failed but fixed that just now so back crunching.

OBVIOUSLY I forgot to reallow virtualisation in Bios so all overnight tasks failed but fixed that just now so back crunching.

Hello Ray,
since the new Boinc-Server upgrade.
In the Details of the Computer you can see now if VT-X or AMD-V is enabled.
Virtualization Virtualbox (5.2.24) installed, CPU has hardware virtualization support and it is enabled.

On my main Linux box linux-e1r2 I have VirtualBox 6.0.2 installed but the LHC BOINC server says no virtualization since the latest kernel upgrade from SuSE. If I click VirtualBox it comes up with no error message.
Tullio
From past experience I know I have to reboot. I have installed VBox 6.0.2 after the kernel upgrade and its consequent reboot.

---

Hi Tullio,
the Details of your Computer say it is enabled, of course.
Had you before 5.2.22 or 5.2.24(newest 5.2.xx) running?

Hi Tullio,
the Details of your Computer say it is enabled, of course.
Had you before 5.2.22 or 5.2.24(newest 5.2.xx) running?

I always upgrade to the latest VBox version by choosing the last option for Linux. It is now running.
Tullio