Message boards :
News :
ATLAS load tests
Message board moderation
| Author | Message |
|---|---|
 Send message Joined: 12 Sep 14 Posts: 42 Credit: 111,031 RAC: 0   |
For information: We have increased limits pr. host on the dev project as part of a campaign to test our new storage backend. Expect most new ATLAS tasks to be pulled by our local cluster hosts and the LCG cluster in Bejing. But there are tasks for other applications too for those interested. There might be interruptions too, as this is our development and testbed. :-) |
 Magic Quantum Mechanic Send message Joined: 8 Apr 15 Posts: 780 Credit: 12,151,079 RAC: 2,173  
|
Well it sure looks like right now you really don't need us @home members since you have these two not@home fleets with more than 100 times the cores I have @home. We have that over at LHC too. Mad Scientist For Life |
|
Send message Joined: 12 Sep 14 Posts: 42 Credit: 111,031 RAC: 0 |
These in-house cores are only available temporarily when there is no analysis jobs going on. So we are more than happy for contributions from all of you! The more capacity we can get, the better :-) Now we recommend to keep connected to the regular LHC@home, as this dev project can be erratic. |
|
Send message Joined: 22 Apr 16 Posts: 677 Credit: 2,002,766 RAC: 565
|
This task is not finished from anonymus and send to my Computer as native(app). https://lhcathomedev.cern.ch/lhcathome-dev/results.php?hostid=2378 There are hundreds of Atlas with Error! |
|
Magic Quantum Mechanic Send message Joined: 8 Apr 15 Posts: 780 Credit: 12,151,079 RAC: 2,173
|
These in-house cores are only available temporarily when there is no analysis jobs going on. So we are more than happy for contributions from all of you! The more capacity we can get, the better :-) Well as you can see I have been here for 3 years running my 10 @home computers and have done more Valid tasks than anyone here yet one of these new fleets you have will pass me up in just TWO months. That seems to me that it is something that will get you LESS of the public here AND over at the LHC ( and I have been over there 24/7 since 2004) and we see the same there. If it was up to me and had those 100's of not@home computers to use here and LHC I sure wouldn't have them listed on the members pages or statistics pages. Some people do like to say that *credits* don't matter but that is the ONLY goal or thing that keeps members here 24/7 for YEARS I may be one of the few that stays here and also reads all the news and updates on official CERN websites that show what the LHC is doing and the CMS,Atlas,and LHCb are actually doing. I also did more than anyone else for over at the previous T4T/vLHC and was the only person there from day one until it was ended and moved to LHC. Members like me pay thousands of dollars to do this (and doing GPU for Einstein where this never has happened and I have been there since the beginning) And THAT is all I do......not a game player who wants credits from EVERY Boinc project. I now start thinking about how I would feel if I was doing this for 15 years 24/7 and then find all I did was test to make it work for the hundreds of computers you have waiting to do this. I also am one of the only 2 members that has been doing the Alpha testing of Atlas. Sure do not hope that the best I can see is my wife telling me about how she always told me for all these years that I am just wasting time and money. But then she thinks Cern and Einstein projects are doing what that Seti project is for.........no matter how many times I tell her there are no alien space crafts or planets sending signals from inside the LHC  (and I retired from seti after the Classic version became a Boinc project and Cern LHC first started.) Mad Scientist For Life |
|
Magic Quantum Mechanic Send message Joined: 8 Apr 15 Posts: 780 Credit: 12,151,079 RAC: 2,173
|
This task is not finished from anonymus and send to my Computer as native(app). Axel, I see you were getting lots of Valids before with ATLAS Simulation v0.50 and then have all those *aborted* tasks. What made you abort them? Where they not starting to run and just saying "waiting to start" or did they say "running" but stay at 0 % process and no CPU running? As you know I don't have any Linux running so your stderr will be different than my Windows version. I did some Atlas here with no problems a couple weeks ago but since I just reinstalled the OS on this pc I would have to d/l this vdi again and it takes several hours to get the Cern server to send that via satellite to my server .......for some reason. But I could try that again and test a few but like I said it would be with a Windows OS (have you tried with a Windows OS lately? ) |
|
Send message Joined: 22 Apr 16 Posts: 677 Credit: 2,002,766 RAC: 565
|
This is a missunderstanding! The hundreds of Error or aborting are from the Anonymus-Member. Have got one of them to run with native-app. This WU is running at the moment. |
 Send message Joined: 28 Jul 16 Posts: 482 Credit: 394,720 RAC: 0 |
The hundreds of Error or aborting are from the Anonymus-Member. The errors are most likely from: nodename=BEIJING_LCG2@lwn145.ihep.ac.cn |
|
Send message Joined: 12 Sep 14 Posts: 42 Credit: 111,031 RAC: 0 |
@MAGIC: The only reason why we direct our temporarily idle hypervisors (Agile Boincers) here is to be able to load-test our new storage back-end with ATLAS load before upgrading our production environment. Under normal circumstances, cluster backfill at CERN is done outside BOINC, as such nodes can handle data-intensive analysis and other work than simulations. Your contributions and notably alpha and beta testing are very much appreciated, and help us to reach more volunteers on a wider range of OS. I suggest to continue the discussion on Windows and ATLAS under the "ATLAS application" forum. |
|
Send message Joined: 10 Mar 17 Posts: 40 Credit: 108,345 RAC: 0 
|
are the tasks in queue just test tasks or are these "real" tasks. i.e. if i enable my pc to run dev project tasks, would this be a "scientific calculation" or would it just be a "calculation/test for the new backend" or both? |
|
Send message Joined: 20 Apr 16 Posts: 180 Credit: 1,355,327 RAC: 0  |
These are real tasks, from the same pool of tasks that go to LHC@Home. So if you run tasks here you are both contributing to the science and testing the new infrastructure. But as Nils said this is still the dev project so things can go wrong at any time! |
|
Send message Joined: 22 Apr 16 Posts: 677 Credit: 2,002,766 RAC: 565
|
Is it possible to get the stats therefore in Boincstats or teamocuk.co.uk. Since weeks is no stats possible for -dev project. Thank you. |
 Send message Joined: 29 May 15 Posts: 147 Credit: 2,842,484 RAC: 0
|
We are getting low on Atlas-WUs, at the moment there are 0 available. For a loadtest of the new backend it would be good to have as much WUs as Volunteers are polling for |
|
Magic Quantum Mechanic Send message Joined: 8 Apr 15 Posts: 780 Credit: 12,151,079 RAC: 2,173
|
Looks like they gave us some more Yeti ( 1882 ) I'm only getting 2 here since it will take hours to d/l over 400MB Mad Scientist For Life |
|
Send message Joined: 29 May 15 Posts: 147 Credit: 2,842,484 RAC: 0
|
We have passed the checkmark of 10.000 WUs out in the field. A little bit more and we will reach the critical number |
|
Send message Joined: 10 Mar 17 Posts: 40 Credit: 108,345 RAC: 0
|
These are real tasks, from the same pool of tasks that go to LHC@Home. So if you run tasks here you are both contributing to the science and testing the new infrastructure. But as Nils said this is still the dev project so things can go wrong at any time! thanks for the info david! pc is now connected to -dev project (just one pc, but better than nothing i guess ;-) ) |
|
Send message Joined: 12 Sep 14 Posts: 42 Credit: 111,031 RAC: 0 |
Test gone well so far. Thanks to you all! We'll have a temporary stop in the supply of tasks to modify the storage back-end. |
©2024 CERN