So, first it was free-dc. Then it was extemeHW (they shorted me for 6 days) now its WCG. What is going on in the DC world ????

[Markfw]

the title says it all.

Edit: just googled and status on x for WCG is down until the morning for maintenance.

 

[Assimilator1]

And Asteroids is out of WUs, LHC 6 track has none (not unusual), nor Universe!
I think this is the 1st time my client has stalled with 4 'active' projects! Guess I need to spread my wings....

Hmm, I have Rosetta set to 'no new tasks', I can't remember why now....
Oh, it doesn't have any WUs anyway

 

[mmonnin03]

And SRBase did some DB/server maintenance which caused all downloads to fail. Admin wants users to work through the bad downloads until the tasks abort 8 times. DB restored from a backup which caused those actively crunching to lose credit.

 

[Assimilator1]

Hmm, I actually don't know what I want to crunch!
Bearing in mind I only crunch with the CPU these days (except for the folding race).
I could crunch Milkway, Einstein or Folding, but these all do best with GPUs IIRC, so not very efficient.
I considered Gia, but no tasks, I considered Uspex but I'm not supporting any Russian based project whilst there government is invading Ukraine!
Evolution is dead I think?
DHEP died ~5 yrs ago (I was suprised to find out this morning! lol. And now the web archive for it is dead too! ).
Muon1 DPAD's finished a long time ago.

I'm not interested in maths projects, but I am in medical and science (especially space related).
Any ideas anyone? I guess in the meantime I'll crunch MW or Einstein, until someone has a better idea, or the others get WUs again.

[edit] Can't run Einstein on CPU only, can't see where to set preferences for that.

 

[Markfw]

WCG is back up ! It wook me about 1/2 hour to play catchup, but the 9654 still has no work. The other 7 are crunching.

 

[Assimilator1]

Must've just happened, lol, my log shows it tried 33 mins ago with no luck.

 

[mmonnin03]

Hmm, I actually don't know what I want to crunch!
Bearing in mind I only crunch with the CPU these days (except for the folding race).
I could crunch Milkway, Einstein or Folding, but these all do best with GPUs IIRC, so not very efficient.
I considered Gia, but no tasks, I considered Uspex but I'm not supporting any Russian based project whilst there government is invading Ukraine!
Evolution is dead I think?
DHEP died ~5 yrs ago (I was suprised to find out this morning! lol. And now the web archive for it is dead too! ).
Muon1 DPAD's finished a long time ago.

I'm not interested in maths projects, but I am in medical and science (especially space related).
Any ideas anyone? I guess in the meantime I'll crunch MW or Einstein, until someone has a better idea, or the others get WUs again.

[edit] Can't run Einstein on CPU only, can't see where to set preferences for that.

E@H: FGRP5 for CPU and has plenty of work
M@H: Is CPU only. Single and MT.

SiDock and LHC pretty much always have work and are not math projects.

 

[Markfw]

So. WCG is sort of up, but no units available.

 

[mmonnin03]

Rosetta has work work. I see tasks using up to 2.5GB of memory.

 

[Assimilator1]

E@H: FGRP5 for CPU and has plenty of work
M@H: Is CPU only. Single and MT.

SiDock and LHC pretty much always have work and are not math projects.

M@H is CPU only now!? Well I didn't know, or had forgotten that! lol
Interesting to know about E@H, just had a look at the preferences to try to set 'resource settings', I see :-
Use CPU: yes/no
Request CPU-only tasks from this project.

Use NVIDIA GPU: yes/no
Request NVIDIA GPU tasks from this project.
But when I click yes or no (they turn the pointer to a hand) nothing happens, any ideas? [edit, Nm, when I looked at the comparison view, it showed the settings had been set after all.


I'm not getting any work for LHC, and very rarely do .
Not familiar with Sidock, I'll have to check that out.

So. WCG is sort of up, but no units available.

Well it's some days later, but I've got plenty of WUs for that one now

 

[mmonnin03]

The Yes/No at E@H changes from between a highlighted Bold to not bold.

At LHC, there is plenty of work for CMS (VBox) Atlas (VBox or Linux Native) Theory (VBox or Linux Native). Theory tasks are currently running for multiple days atm. Whenever they have SixTrack tasks they run out very quickly.

 

[StefanR5R]

I'm not getting any work for LHC, and very rarely do .

At LHC, there is plenty of work for CMS (VBox) Atlas (VBox or Linux Native) Theory (VBox or Linux Native). Theory tasks are currently running for multiple days atm. Whenever they have SixTrack tasks they run out very quickly.

@Assimilator1, in order to receive "Linux native" work, you obviously need Linux, plus various extra stuff to set up, basically related to CERN's distributed filesystem. In order to receive VirtualBox based work, you obviously need to have VirtualBox ready to roll on the box and recognized by BOINC.

On top of that, the host which is asking for work also needs to have a good deal of free RAM and free disk space. Below a certain threshold, the server won't hand out work. How much of that exactly is hopefully documented somewhere at LHC@home's message board. Furthermore, at least ATLAS work is multithreaded. Therefore it _may_ be that hosts which ask for work need to have more than 1 logical CPU usable for BOINC in order to receive any work. But I am not sure of that, perhaps just 1 active logical CPU on a host gets work assigned too.

M@H is CPU only now!? Well I didn't know, or had forgotten that! lol

Plus, most of the time (but not always) they have work ready to assign.

BTW, @Kiska's server status tracker: LHC@home grafana, MilkyWay@Home grafana

 

[Ken g6]

LODA's server is more or less out of disk space, so I can't upload results.

 

[Assimilator1]

Who?

The Yes/No at E@H changes from between a highlighted Bold to not bold.

At LHC, there is plenty of work for CMS (VBox) Atlas (VBox or Linux Native) Theory (VBox or Linux Native). Theory tasks are currently running for multiple days atm. Whenever they have SixTrack tasks they run out very quickly.

Re E@H, ah yea, I see now, it's barely visible unless I zoom right in! lol, what a rubbish design!

Re LHC, ah yea I remember now, I removed Vbox, I think I stopped using it as their was either a lack of WUs, a problem with them, or a problem with Vbox a few years ago or so. After that I never re-installed it as I didn't have a need for it. So I think that only left me with the 6 track which has very few WUs, a shame really.

@StefanR5R as above I no longer have Vbox, I can't really remember what the problem was now, although I do recall posting here about it (I think, can't remember where![edit] Found it - https://lhcathome.cern.ch/lhcathome/forum_thread.php?id=4304&postid=46284 ), I don't think I was the only one though. Perhaps I should re-visit Vbox....
My main rig has plenty of RAM (specs below), I would've thought 32GB would be enough? CPU is 6C/12T.

Atm I'm getting plenty of WUs (now) from Asteroids, WCG, Einstein & MW , so plenty of WUs but not enough CPU time! lol

 

[StefanR5R]

Actually I have been avoiding LHC@Home's VirtualBox based applications myself for maybe years now, for about the same reasons you did. In addition, I dislike that virtual machine based applications circumvent BOINC's CPU/RAM/disk/network resource control facilities. I did run Cosmology@Home's VirtualBox based application for quite some time which exhibited a mild form of the "postponed" tasks phenomenon. I have read advice (at I don't recall which message board) that manual replacement of the vboxwrapper binary by a newer version could help with that, but I never tested this. Instead I merely used a little background script which periodically cleaned out any "postponed" Cosmology tasks.

 

[mmonnin03]

Until last night when LHC ran out of CMS tasks, I've been running that sub project which is VBox only. I have it set to 4 threads per task which uses around 4.5GB of memory so well within the 2GB of RAM/thread I have on each PC. Credit was pretty decent running between 7-8 tasks at once on a 7950x at a bit over 110k/day. The later versions of VBox/BOINC do seem better with the postponed situation. I haven't had any of that in awhile.

 

[Ken g6]

LODA's server is more or less out of disk space, so I can't upload results.

LODA had to load-a from a backup. Credits were lost.