Message boards :
News :
Progress!
Message board moderation
Author | Message |
---|---|
Send message Joined: 12 Sep 14 Posts: 1069 Credit: 334,882 RAC: 0 |
We are making great progress and are just chasing up the last remaining issues. One of the recent improvements was to create the link to the CMS monitoring infrastructure. This means that we can generate nice plots similar to what ATLAS@home have for their project. |
Send message Joined: 1 Aug 14 Posts: 14 Credit: 884 RAC: 0 |
That's great! I can't wait to see those plots :) Especially those that compare CMS@Home to the "traditional" CMS resources. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 |
Looks great. What is the unit on the Y-axis? |
Send message Joined: 13 Feb 15 Posts: 1188 Credit: 861,475 RAC: 2 |
Looks great. Looks like received jobs/hour. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 |
Thanks. Are these the "jobs"? Begin processing the 1st record. Run 1, Event 1266, LumiSection 254 at 28-Aug-2015 04:00:09.508 CEST |
Send message Joined: 13 Feb 15 Posts: 1188 Credit: 861,475 RAC: 2 |
Thanks. No. A job is 1 run of the cmsRun process. 1 job/cmsRun is stored in your http://localhost:?????/logs/run-?/glide_??????/dir_????/ Fill in the right figures for the question marks. Ivan can configure how many records (in CERN they're speaking about events) he wants to fit in one cmsRun. Last week we had jobs with 200 events (records). |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
At the moment I'm using 5 events/job to get quick turnaround for Laurence and Hassen to debug. Expect that to increase tonight. :-) Basically' I'm looking for jobs that run in 30-60 minutes and create 20-30 MB output each. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 |
Thank you. So, the number of "records" per run need to be the same, otherwise the "jobs" are not comparable in the statistics. |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
Thank you. Yes. There other metrics, like the wall-clock time, etc. |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
OK, the last test was a great success, so I'll dump a load of larger jobs in a little while, I'm just doing a pilot run on the GRID proper before starting the test. Thanks for your help, in the past and in the future. ivan |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
Unfortunately the next lot of jobs will be a bit later than planned -- the CRAB server was upgraded this afternoon, which broke "our" patch. I'm told it will be fixed in about two hours, so I'll try again then. |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
Unfortunately the next lot of jobs will be a bit later than planned -- the CRAB server was upgraded this afternoon, which broke "our" patch. I'm told it will be fixed in about two hours, so I'll try again then. Not yet... |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 |
Is there going to be any work over the weekend? If not, i am going to shut it down till monday. |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
Looks like it's not going to be fixed tonight. Sorry... Meanwhile, 1969 of the supposed parallel jobs submitted to the GRID have finished, with 30 errors and one still to finish... :-( |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
Is there going to be any work over the weekend? It's not looking good, I'm afraid. Feel free to switch off. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 |
Thanks for letting me know. |
Send message Joined: 20 Mar 15 Posts: 243 Credit: 886,442 RAC: 0 |
Unfortunately the next lot of jobs will be a bit later than planned -- the CRAB server was upgraded this afternoon, which broke "our" patch. I'm told it will be fixed in about two hours,... Oh dear, oh dear, oh dear. Bad move, that. Thanks for the info. |
Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 728 |
Sorry to say that this isn't resolved yet. In fact, the problem has raised some "differences of perspective" amongst the leading players. I'm staying out of that as far as possible, I'm as much a volunteer as you lot. But there's hope we can get it sorted soon. It's a long weekend here, so I'll be buying lots of popcorn to watch tomorrow's round of emails... |
Send message Joined: 29 May 15 Posts: 147 Credit: 2,842,484 RAC: 0 |
This remembers me back to the early years of LHC@Home and sixtrack; it must have been similar ;-) |
©2024 CERN