Message boards :
Theory Application :
CPU non used
Message board moderation
Previous · 1 · 2 · 3
Author | Message |
---|---|
Send message Joined: 15 Jun 08 Posts: 2402 Credit: 225,611,152 RAC: 121,525 |
I am running BOINC as root So missing write access should not be the problem. Although I do not recommend to run BOINC under the root account. connection to vocms0159.cern.ch 4080 port [tcp.lorica-in] succeeded! Contact to WMAgent is also possible. What makes me pause is the following line in your log: Processor supports HW virtualization: no You may do some additional tests regarding the VM settings on your machine. 1. Deinstall other virtualization software, e.g QEMU, KVM, XEN, WINE. 2. Check if AMD-V is enabled in the BIOS 3. Check position 5. of Yeti´s checklist 4. run "grep -i ' svm ' /proc/cpuinfo". The output must contain 1 line per CPU core. 5a. Install the package "msr-tools" 5b. run "modprobe msr" 5c. run "rdmsr 0xc0010114" and post the result here. This reads out the VM_CR register of the CPU where VMs can be blocked. |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
AMD-V is enabled in BIOS, also KVM. Xen and Wine are not installed. I am now running an Atlas double core from LHC and it seems right. Thank you for your support. Tullio flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow rep_good nopl extd_apicid pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy 3dnowprefetch vmmcall flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt rdtscp lm 3dnowext 3dnow rep_good nopl extd_apicid pni cx16 lahf_lm cmp_legacy svm extapic cr8_legacy 3dnowprefetch vmmcall 0 |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
Atlas 1.01 completed and validated. Tullio |
Send message Joined: 15 Jun 08 Posts: 2402 Credit: 225,611,152 RAC: 121,525 |
AMD-V is enabled in BIOS, also KVM. Xen and Wine are not installed. I am now running an Atlas double core from LHC and it seems right. Thank you for your support. Everything looks good. |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
It's a good old SUN WS running since 2008 on SuSE Linux. I had only to change hard disks and I added a GTX 750 GPU board for SETI and SETI Beta. It has not enough RAM for Einstein GPUs, which run on my Windows 10 with a GTX 1050. It's a wonder what a GPU board can so. Tullio |
Send message Joined: 15 Jun 08 Posts: 2402 Credit: 225,611,152 RAC: 121,525 |
Atlas 1.01 completed and validated. Looks good at first sight as the WU finished successfully. At second sight, hm? A 2 core WU with a runtime of 8000 s should have a CPU time close to 16000 s instead of 2341 s. The logs should be checked by the ATLAS experts. |
Send message Joined: 15 Jun 08 Posts: 2402 Credit: 225,611,152 RAC: 121,525 |
It's a good old SUN WS running since 2008 on SuSE Linux. I had only to change hard disks and I added a GTX 750 GPU board for SETI and SETI Beta. ... As it is a 2 core host I would keep an eye at the avg_load values. They should not go much over "2" for more than a few seconds. The higher the shorter. To keep the value low I would configure the CERN VMs to use only 1 core and let the remaining core free for the OS and to feed the graphic card tasks. |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
I am using the "top" command to watch the CPU usage. Tullio |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
Atlas 1.01 completed and validated. I see an Atlas 1.01 task completed but not validated on the Windows 10 PC with 1,111.64 s elapsed time and 489.69 s CPU time, yet it is multicore. Tullio |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
Two more Atlas tasks completed and validated on the Linux box while those on the Windows 10 PC are not validated. All mtcore tasks. Tullio |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
I have a Theory task running alone on my Windows 10 PC which is not doing anything, CPU usage is close to zero, yet the BOINC task is going on hour after hour, using my CPU and my electricity. Can anybody explain this to me? Tullio |
Send message Joined: 14 Jan 10 Posts: 1276 Credit: 8,481,858 RAC: 1,977 |
I have a Theory task running alone on my Windows 10 PC which is not doing anything, CPU usage is close to zero, yet the BOINC task is going on hour after hour, using my CPU and my electricity. Can anybody explain this to me? No, I cannot. Last week also a CMS-task and a few ATLAS-tasks fail on that machine. Have you had a look into the VM-logs or the Remote Desktop? |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
I've looked the logs but I am not able to understand them. Maybe someone from CERN should have a look at them. I have been running Test4Theory@home tasks since November 2010 after an invitation by dr.Segal and my ID number was 10. Now that VLHC@home has disappeared all this is just a memory. Tullio |
Send message Joined: 19 Feb 08 Posts: 708 Credit: 4,336,250 RAC: 0 |
SUSE have today a new Vitualbox and a reinstall from Boinc. Yes,true. I installed SuSE Leap 42.2 64-bit on the HP Laptop with APU E-450 using a USB stick. But LHC tasks all fail, except SixTrack. Tullio |
Send message Joined: 22 Nov 05 Posts: 8 Credit: 136,465 RAC: 0 |
Theory Simulation 262.70 (vbox64) : Status -> Verschoben:VM Hypervisor failed to enter an online state in a timely fashion. ATLAS Simulation 1.01 (vbox64_mt_mcore_atlas) : State -> Verschoben:VM Hypervisor failed to enter an online state in a timely fashion. (8 CPUs) |
Send message Joined: 14 Jan 10 Posts: 1276 Credit: 8,481,858 RAC: 1,977 |
Theory Simulation 262.70 (vbox64) : Status -> Verschoben:VM Hypervisor failed to enter an online state in a timely fashion. Mostly a restart of the BOINC client will restart those tasks, however when the reason for their "waiting to run" is still present after some time the task will wait again for a whole day. Then you could have a look with VirtualBox Media Manager (via VirtualBox Manager -> File), whether there are VM-disks in an error state. |
Send message Joined: 2 May 07 Posts: 2090 Credit: 158,976,078 RAC: 124,119 |
This WU was running more than 1 hour idle after the last work was done: https://lhcathome.cern.ch/lhcathome/result.php?resultid=178327125 2018-02-17 19:24:53 (10764): Status Report: Job Duration: '64800.000000' 2018-02-17 19:24:53 (10764): Status Report: Elapsed Time: '60004.952793' 2018-02-17 19:24:53 (10764): Status Report: CPU Time: '57237.749706' 2018-02-17 20:47:28 (10764): Powering off VM. 2018-02-17 20:52:36 (10764): VM did not power off when requested. 2018-02-17 20:52:36 (10764): VM was successfully terminated. 2018-02-17 20:52:36 (10764): Deregistering VM. (boinc_896865f0a8d291f6, slot#0) vm_image.vdi entry is not deleted! |
©2024 CERN