Message boards :
CMS Application :
CMS Simulation v60.50
Message board moderation
Author | Message |
---|---|
Send message Joined: 13 Feb 15 Posts: 1188 Credit: 861,827 RAC: 35 |
lhcathome-dev 08 Jul 12:48:02 Started download of CMS_2021_07_07.vdi lhcathome-dev 08 Jul 12:54:33 Finished download of CMS_2021_07_07.vdi I'm not using a proxy. In version 60.40 http://128.142.248.156:3125 and http://128.142.161.84:3125 was used, now DIRECT. Why is DIRACbenchmark. running during startup? Is this benchmark really used for something? |
Send message Joined: 28 Jul 16 Posts: 484 Credit: 394,839 RAC: 0 |
Version 60.40 was configured to use the traditional stratum-1-servers for CVMFS. If no local proxy is available this automatically configures proxies at CERN/Fermilab to avoid going DIRECT to the stratum-1s. The reason for that were 2 missing files (on the vdi) from the recent cvmfs-config-default package. Version 60.50 includes those files and now the VM's CVMFS correctly configures openhtc.io which is always contacted DIRECT (or via a local proxy set by the volunteer). DIRACbenchmark runs on all CMS vdis for years but I also don't know why. |
Send message Joined: 13 Feb 15 Posts: 1188 Credit: 861,827 RAC: 35 |
DIRACbenchmark runs on all CMS vdis for years but I also don't know why.I never saw DIRACbenchmark as a process before. Probably cause it was done before one could use Console F3. The benchmark was done before during startup and the outcome was shown on Console F1, but is not displayed anymore. In my opinion that benchmark is useless. The outcome differs very much, depending whether (all) other cores of the host are idle or crunching other tasks. Could it be skipped? |
Send message Joined: 28 Jul 16 Posts: 484 Credit: 394,839 RAC: 0 |
Your experience is the same that I made. I'd like to support your request to skip that benchmark. Or can anybody from the project team explain why it is a must? |
Send message Joined: 8 Apr 15 Posts: 781 Credit: 12,374,481 RAC: 4,766 |
7/8/2021 3:23:25 PM | lhcathome-dev | Started download of CMS_2021_07_07.vdi ( 1.60GB) Of course mine will take longer than an hour to d/l so I won't post the finish time (326KBps) here |
Send message Joined: 8 Apr 15 Posts: 781 Credit: 12,374,481 RAC: 4,766 |
7/8/2021 3:23:25 PM | lhcathome-dev | Started download of CMS_2021_07_07.vdi 7/8/2021 4:42:27 PM | lhcathome-dev | Finished download of CMS_2021_07_07.vdi Never have seen this before Oasis.openscience.org |
Send message Joined: 22 Apr 16 Posts: 677 Credit: 2,002,766 RAC: 1 |
https://opensciencegrid.org/ Our Cores are peanuts ;-) |
Send message Joined: 28 Jul 16 Posts: 484 Credit: 394,839 RAC: 0 |
Never have seen this before ... Each CMS task since March 2019 downloads data from /cvmfs/oasis.opensciencegrid.org. I'ts just not mentioned in the logfiles. Now bootstrap-cms has been changed to mention it. |
Send message Joined: 8 Apr 15 Posts: 781 Credit: 12,374,481 RAC: 4,766 |
https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3000633 Guess I will switch back to Sixtracks for a while |
Send message Joined: 8 Apr 15 Posts: 781 Credit: 12,374,481 RAC: 4,766 |
7/22/2021 11:57:25 AM | lhcathome-dev | Started download of CMS_2021_07_07.vdi I guess I should d/l the 1.6GB .vdi on this machine now .....looks like it is running at 350KBps+ right now so that should be fairly fast for my Dish and that way I can finish the Sixtracks aka McIntoshTracks |
Send message Joined: 8 Apr 15 Posts: 781 Credit: 12,374,481 RAC: 4,766 |
Yes, nothing like an *anonymous" computer here. https://lhcathomedev.cern.ch/lhcathome-dev/results.php?hostid=351 I think I know who it is anyway |
©2024 CERN