| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: BOINC queue clobbered. Why? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 3
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Short version:
Running Fedora Linux, I ran boincmgr while WCG was suspended and boinc-client.service was stopped. This appears to have "reset" WCG, clobbering all work units. Does that sound correct? Long version: Been having problems with my wifi hotspot causing hard freezes, which happened this morning. After that happended, and I rebooted, I did a suspend on WCG (stopping boinc-client.service), intending to do some stuff then reboot, but then didn't reboot, and then went into boinc manager, which didn't show anything (boinc-client.service was stopped), and asked me to pick a project, but I cancelled out and then did a start boinc-client.service (as root). At that point, starting boincmgr again, showed my boinc queue had been cleared, so I selected WCG. I'm guessing that starting boincmgr, while boinc-client.service wasn't running, clobbered everything. It looks like boinc is recovering nicely - redownloading the wus. I've had several hard freezes recently (kernel / wifi problem), and that never bothered boinc before. A whole bunch of work units now show a status of "Detached" on the WCG website. In boincmgr, about half of the new work units show "download failed", while 4 are running (as usual), and a couple are ready to start. Waiting to see if boinc eventually recovers before worrying about next steps. Any advice? |
||
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2351 Status: Offline Project Badges: 
|
I'm running Fedora Linux, too.
I think this is what happened: Normally your queue is in ~boinc. Then, when boinc-client.service is running, you can connect to the service and it will find the WUs in ~boinc. However, when boinc-client.service isn't running, boincmgr will try and set up a private instance of all BOINC-files in your $HOME and run there. Somehow things got mixed up and WCG was thinking that you lost your queue and were needing re-creation of your BOINC-queue, which it did obediently. Check your $HOME to see if any BOINC-related files and directories are created there and if so, remove them. BOINC should normally be running in ~boinc. Additional files that are created are "$HOME/.Boinc Manager" and the directory "$HOME/.BOINC/". |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Right you are. I did notice a couple of boinc files in my home directory that I'd never noticed before. And now, with your explanation, I see there are about a dozen or so. And the time stamps of the home directory files are all around noon, whereas the time stamps of the files in /var/lib/boinc are roughly now (5 PM). So I will delete the $HOME files and keep my fingers crossed. Wow, remind me never to do that again! Thanks for making sense of it.
|
||
|
|
|