Message boards :
CMS Application :
no new WUs available
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 24 · Next
| Author | Message |
|---|---|
 Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261  |
According to the Project Status Page, CMS has run out of tasks :-( Yes, I'm trying to work out what's the cause. Trouble trying to contact Laurence, of course. We are still serving up jobs over at -dev if you are really desperate for heating power from your PCs. :-)  |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794  
|
... if you are really desperate for heating power from your PCs. :-)haha, right now we have springtime temperatures here in Austria. Merry X-Mas :-) |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
... if you are really desperate for heating power from your PCs. :-)haha, right now we have springtime temperatures here in Austria. Tja! Nice and mild here. My gas boiler is bjorked so I'm relying on expensive electrickery to keep warm. Minimum temperatures close to double-digits are keeping my costs down! |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
According to the Project Status Page, CMS has run out of tasks :-( OK, I'm not sure what broke the log-jam (no messages in my in-box) but our task creator seems to be working again -- 186 pending tasks in the queue. I logged on to my Uni cruncher just now and persuaded it to download two tasks, so it looks like we're in business again. Have a warm Christmas (unless you're back home in Byron Bay, then rather search out the air-conditioner or de-humidifier!). |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794
|
hello Ivan, the queue ran dry :-( will there be new tasks soon? |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
hello Ivan, the queue ran dry :-( Hi; I'm not sure if the server status page is misleading or not. There seems to be little change to the actual jobs running. I'll do some further checks and alert the admins. [Edit] OK, I'm running fewer tasks than I should, so task creation is stopped. WMagent is fine and jobs are being issued to running tasks. I'll e-mail the crew... [/Edit] |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
Running job numbers are falling as tasks complete and new ones aren't served. This is a bit more benign than having tasks but no jobs, but set No New Tasks anyway if you wish. |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794
|
... new ones aren't served.:-( |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
No response yet, guess everybody's working flat out on patching the Meltdown vulnerability: http://www.theregister.co.uk/2018/01/04/intel_amd_arm_cpu_vulnerability/. Looks like we'll all have to download patched VMs, too. ...and I'll be busy patching my array of servers just as soon as the patches arrive... |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794
|
...and I'll be busy patching my array of servers just as soon as the patches arrive...the bad news is that this patch will cause all Intel processors to slow down considerably :-( |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
...and I'll be busy patching my array of servers just as soon as the patches arrive...the bad news is that this patch will cause all Intel processors to slow down considerably :-( One of my cow-orkers says that the UK LHC grid community is considering specifying AMD only from now on (worryingly, the Register article suggests they are also vulnerable to Spectre, if not Meltdown). This is a problem as we have a tender out at the moment for kit that needs to be delivered by the end of March to be paid by our grant -- "Use it or lose it!". |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
There's a hint we may have tasks again. My jobs monitor is showing an upwards tick. The publicly-accessible server status page and job status plots are only updated hourly, so it's too soon to be sure. I'm off to work now, more later as the story unfolds... |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
Looks like a false alarm, just a blip on the graphs. I need to check what CERN's Christmas/New Year closure dates are... [Edit] Oops, looks like CERN is closed until Monday! [/Edit] |
|
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 
|
Maybe the Three Magis (Wizards?) will bring some tasks on the Epiphany. Their bones were in Sant'Eustorgio Church in Milan, but a German Emperor abducted them to Koln. Tullio |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794
|
Looks like a false alarm, just a blip on the graphs. I need to check what CERN's Christmas/New Year closure dates are...Ivan, how come that the Project Status Page still shows 0 unsent tasks, and the dashboard of running jobs (https://lhcathomedev.cern.ch/lhcathome-dev/cms_job.php) still shows a continuous number of o/a 1000 tasks? |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
Looks like a false alarm, just a blip on the graphs. I need to check what CERN's Christmas/New Year closure dates are...Ivan, how come that the Project Status Page still shows 0 unsent tasks, and the dashboard of running jobs (https://lhcathomedev.cern.ch/lhcathome-dev/cms_job.php) still shows a continuous number of o/a 1000 tasks? Yes, I thought someone would ask that... Laurence maintains a farm of VMs at CERN that runs jobs that volunteer machines cannot -- basically anything beyond the first-stage Monte Carlo simulations, but especially the "merge" jobs that combine the ~60 MB result files into ~2.5 GB final files to ease the problems of file distribution and handling. But, they run anything the HTCondor server dispatches, the caveat being that the latter-stage jobs are only presented to the local CERN machines -- they will still run "normal" jobs if they are available. Laurence has been slightly coy to me about how many VMs he actually controls -- they all come out of a central CERN allocation to CMS that we are not otherwise using -- but now we know that it's currently about 1,000! :-) |
|
Send message Joined: 24 Jul 16 Posts: 88 Credit: 239,917 RAC: 0 
|
I don't know if there is a link with this opportunistic farm of computers and a discrepancy noticed between informations displayed on project status and application page. On status project : Current GigaFLOPS 59378.42 Task data as of 7 Jan 2018, 11:04:59 UTC on application page : Total average computing: 48,523 GigaFLOPS We deduce a difference of nearly 11 000 GigaFLOPS (variable from 9 000 to 11 000 , since 2 weeks ago~, average 10 000 GigaFlops) Does it mean that the opportunistic farm has this powerfullness ? By the way , it would be good to add lines written "other" in each sub project in application page to include "platform unknown or other platform" to avoid this discrepancy in GigaFLOPS. |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794
|
Ivan, any chance for new CMS tasks soon? |
|
Send message Joined: 29 Aug 05 Posts: 1065 Credit: 7,884,757 RAC: 11,261 |
Ivan, any chance for new CMS tasks soon? There appear to be some now! I'm still waking up so it will take a few minutes to be sure but the server status page says tasks are available, and my monitors are starting to inch upwards. [Edit] Yep! I got new tasks at 0754, looks like CERN is back from holidays. [/Edit] |
|
Send message Joined: 18 Dec 15 Posts: 1827 Credit: 119,535,315 RAC: 42,794
|
Ivan, are you aware that no CMS tasks are available? |
©2025 CERN