Questions and Answers :
Wish list :
Send me email on failures
Message board moderation
Author | Message |
---|---|
Send message Joined: 18 Mar 05 Posts: 2 Credit: 288,744 RAC: 0 |
I don\'t actively monitor every system very closely, so I need the system to send me an email when failures occur, such as \"client error\" status on a work unit. |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Hi Larry I can see why a notification email would be useful to you, but I can\'t think of any way to make boinc do it. I\'m not sure whether this would be feasible or not. I wonder what would happen if some projects wanted the feature but some didn\'t. I know that a number of members have workunits running on computers they don\'t see every day, or even every week. What do others think? Cpdn news |
Send message Joined: 5 Feb 05 Posts: 465 Credit: 1,914,189 RAC: 0 |
BOINC itself is pretty programmable, if the features are there. You can turn on and off most of the features you want. So, this would be a BOINC reprogramming issue. There is a BOINC maillist, and forums that are good for these types of suggestions. |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
Hi PoohBear If after a couple of days nobody states any major objections here, I may post the idea on the boinc_dev forum. If nobody there comes up with big objections either, I could complete a Boinc Trac ticket to see what the developers think. I think we\'re now being encouraged to make new suggestions through Trac tickets rather than through the mailing list. Maybe it\'s Kathryn who then allocates each ticket to a developer. I\'m sure some of the \'absentee farmers\' would welcome the chance to have a bit more control over WU progress on computers that are physically remote. Perhaps this matters more for members running long WUs like cpdn. Cpdn news |
Send message Joined: 5 Feb 05 Posts: 465 Credit: 1,914,189 RAC: 0 |
My suggestion would be that it goes in your private box on BOINC (which is available already in new version, if you log into SETI). As they fix that, people can get notifications to their personal e-mail box, and read the information online, where it does them the most good. |
Send message Joined: 29 Sep 04 Posts: 2363 Credit: 14,611,758 RAC: 0 |
I think you mean the posters\' private messages box on the project forum? The forum software here already supports it as far as I know, but Milo\'s waiting to see how discussions about boinc PMs pan out on other project forums. Rytis is still perfecting at least one add-on for the facility, so I imagine that Milo will wait a while before deciding whether to implement PMs here. (I hope I haven\'t let any cats out of bags here, but it must by now be an open secret that the boinc forums are, one after another, implementing a new rather rudimentary PM facility. The much more sophisticated PM facility on the independent cpdn forum has of course always been available to all of us. Find it through my sig.) Cpdn news |
Send message Joined: 8 Oct 05 Posts: 36 Credit: 299,711 RAC: 0 |
With the boinccmd tool, it is posible to connect to the BOINC client and verify the status of the WUs and running tasks, etc. So I think, with a little shell skript, it is posible to monitor the BOINC client, the system, and the CPU utilistation. But a monitor feature in the BOINC client would also be nice. |
Send message Joined: 15 May 09 Posts: 4346 Credit: 16,541,921 RAC: 6,087 |
Anything come of this? When I have looked to see what is happening with my tasks and checking what happens with other tasks from the same work unit I have found many computers running CPDN wher every single work unit fails. Is it possible to set things up so that something pops up in the notices tab of BOINC manager to say for instance, "Your last three tasks have all crashed, please check out your system." Long term I would like to see it post a link to likely causes and fixes but I think that is being overly optimistic. Seriously though, the number of computers downloading models that crash without doing any work is a terrible waste of resources and a message through the notices tab in BOINC could possibly save some of this. Dave |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
This mass crashing is well known to us, and there's a manual procedure to deal with it. But with the other project problems, it's not been used much lately. Backups: Here |
©2024 climateprediction.net