Message boards :
Number crunching :
Server problem?
Message board moderation
| Author | Message |
|---|---|
 Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268   
|
My win11 host has started to see server problems but my win10 host is still connecting OK and getting new work. I did a project reset on problem host but the problem persists. Here is what I see on the message log: 86 LHC@home 17-09-2025 20:54 [sched_op] Starting scheduler request 87 LHC@home 17-09-2025 20:54 Sending scheduler request: To fetch work. 88 LHC@home 17-09-2025 20:54 Requesting new tasks for CPU and NVIDIA GPU 89 LHC@home 17-09-2025 20:54 [sched_op] CPU work request: 95040.00 seconds; 0.00 devices 90 LHC@home 17-09-2025 20:54 [sched_op] NVIDIA GPU work request: 1001.00 seconds; 0.00 devices 91 LHC@home 17-09-2025 20:54 [sched_op] AMD/ATI GPU work request: 0.00 seconds; 0.00 devices 92 LHC@home 17-09-2025 20:54 Scheduler request completed: got 0 new tasks 93 LHC@home 17-09-2025 20:54 Server can't open log file (../log_boinc01/scheduler.log) 94 LHC@home 17-09-2025 20:54 [sched_op] Deferring communication for 00:03:34 95 LHC@home 17-09-2025 20:54 [sched_op] Reason: project is down This host has BOINC 8.0.2. So the problem seems to be Server can't open log file (../log_boinc01/scheduler.log) but how to resolve this situation? [edit] The same message appeared also before project reset. [edit2] The situation started when I aborted a Theory task that had gone past the deadline and server had marked it: Timed out - no response. Here is the host: https://lhcathome.cern.ch/lhcathome/show_host_detail.php?hostid=10858236 I am not sure if the project reset went through as the host shows still that it has all the same tasks as before. [edit3] I will try to detach and reattach the host . 
|
|
Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268
|
Removing LHC@home and adding it again didn't help. Same error appears: Server can't open log file (../log_boinc01/scheduler.log)
|
|
Send message Joined: 17 Aug 17 Posts: 124 Credit: 11,758,907 RAC: 44,085
|
Getting the same error here as well |
 Magic Quantum Mechanic Send message Joined: 24 Oct 04 Posts: 1242 Credit: 83,903,627 RAC: 130,123 
|
9/17/2025 12:18:06 PM | LHC@home | Server can't open log file (../log_boinc01/scheduler.log) Same here |
|
Send message Joined: 6 Feb 23 Posts: 7 Credit: 170,165 RAC: 61 |
This is server side, not client side, issue is with LHC server. Paul. |
|
Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268
|
Oddly the win10 host (BOINC 7.16.5) is still communicating normally with server and getting new Atlas work. The problem host didn't get even a host id from server after adding LHC as a new project. Boinc Manager shows host id as 0.
|
|
Send message Joined: 17 Oct 06 Posts: 94 Credit: 62,593,040 RAC: 45,752
|
I was able to fix this by going out to my boinc data folder and just making a file called scheduler.log Edit: Nevermind I just realized that error message was coming from the server and I just got lucky when I tried that . |
|
Magic Quantum Mechanic Send message Joined: 24 Oct 04 Posts: 1242 Credit: 83,903,627 RAC: 130,123
|
I can't speak for everyone but the reason I mentioned it was so that people on the server side would see us saying this and then they would do their part to take care of that. If we don't do that it doesn't get fixed and next thing it is a weekend and nobody looks at anything until the next monday at best. |
 Send message Joined: 7 Aug 11 Posts: 119 Credit: 30,698,788 RAC: 76,744 
|
Thu 18 Sep 2025 13:53:37 | LHC@home | update requested by user Thu 18 Sep 2025 13:53:41 | LHC@home | Sending scheduler request: Requested by user. Thu 18 Sep 2025 13:53:41 | LHC@home | Requesting new tasks for CPU Thu 18 Sep 2025 13:53:42 | LHC@home | Scheduler request completed: got 0 new tasks Thu 18 Sep 2025 13:53:42 | LHC@home | Server can't open log file (../log_boinc01/scheduler.log) Still happening here |
 Send message Joined: 15 Jun 08 Posts: 2703 Credit: 290,788,048 RAC: 141,359 |
This is a server side issue on boinc01.cern.ch. So far boinc02.cern.ch is not affected. It can't be solved on the client side without tweaking the own DNS since load balancing between both is done via DNS, hence by random. It's not worth to do so. Admins are informed. |
|
Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268
|
For me the problem was solved about an hour and a half ago. Crunching is going on and cache is full :-) The nice people at Cern got it going, I didn't do anything. Thank you!
|
|
Send message Joined: 5 Apr 25 Posts: 56 Credit: 1,098,480 RAC: 4,070   
|
Any info on why all servers besides scheduler, upload and download are offline? 
|
|
Send message Joined: 18 Dec 15 Posts: 1921 Credit: 148,162,664 RAC: 132,304
|
Any info on why all servers besides scheduler, upload and download are offline?Here CMS tasks were uploaded, but could not get reported. When pushing the Update button on the left hand side of the BOINC manager, in the status column it says "communication deferred - and the time counts downwards from 1 hour). The BOINC event log says "Project server temporarily down for maintenance" and "project requested delay of 3600 seconds". Also, the number of unsent CMS tasks on the server status page is shrinking. |
|
Send message Joined: 5 Apr 25 Posts: 56 Credit: 1,098,480 RAC: 4,070
|
All systems seem to be working now, validation queue is gone.
|
|
Send message Joined: 18 Dec 15 Posts: 1921 Credit: 148,162,664 RAC: 132,304
|
All systems seem to be working now, validation queue is gone.here, too :-) |
|
Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268
|
I'm getting now message: LHC@home 13-11-2025 14:33 Server can't parse configuration file And Theory uploads are stuck wtih error: LHC@home 13-11-2025 14:36 [error] Error reported by file upload server: can't parse config file
|
|
Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268
|
I'm getting now message: LHC@home 13-11-2025 14:33 Server can't parse configuration file Server error has cleared and uploads are fine but no Theory tasks are downloading. Server just reports that there are no tasks for Theory, Atlas, sixtrack and xtrack although Theory tasks are available according to server status page. I haven't tried to get CMS just yet.
|
|
Send message Joined: 27 Sep 08 Posts: 888 Credit: 757,926,804 RAC: 349,774  |
I think CMS is the same, I ask for all project and am getting 0 returned. |
|
Send message Joined: 28 Sep 04 Posts: 790 Credit: 61,351,248 RAC: 52,268
|
Now I see on one host that it is getting only 1 Theory task at a time. And there is only 1 task waiting for crunching. So number of tasks in progress is max_concurrent (from app_config) + 1. The other host still has 4 tasks in cache that was downloaded yesterday. Soon I'll see if this host starts to show same behavior.
|
|
Send message Joined: 14 Jan 10 Posts: 1468 Credit: 9,913,839 RAC: 2,085  |
Now I see on one host that it is getting only 1 Theory task at a time. And there is only 1 task waiting for crunching. So number of tasks in progress is max_concurrent (from app_config) + 1.Do you have set the number 1 in the prefs for Max # CPUs? |
©2025 CERN