Message boards : Number crunching : NO WORK!
Message board moderation
Author | Message |
---|---|
Send message Joined: 31 Dec 07 Posts: 1152 Credit: 22,363,583 RAC: 5,022 |
In the several years that I have been running this project I have never before seen a time when there was not work available. Does anyone have any idea when there will be new WU’s available for anything? |
Send message Joined: 5 Aug 04 Posts: 1496 Credit: 95,522,203 RAC: 0 |
Nothing specific. We were advised that new work would be posted when the queue dropped below a thousand tasks. Well, we're a thousand below a thousand tasks now, so we're due. There is a batch of FAMOUS work on the shelf because of server storage issues. That might be true for for other research projects. As soon as Mods know, rest assured that you'll know. "We have met the enemy and he is us." -- Pogo Greetings from coastal Washington state, the scenic US Pacific Northwest. |
Send message Joined: 11 May 07 Posts: 36 Credit: 1,485,638 RAC: 0 |
I have the same issue - no work! Its the first time I've seen it - No comment on when work will be available from the moderators...? GH |
Send message Joined: 15 May 09 Posts: 4540 Credit: 19,039,635 RAC: 18,944 |
There has been a post in the news and anouncements section saying they hope to have more work within the next couple of days. |
Send message Joined: 5 Aug 04 Posts: 1496 Credit: 95,522,203 RAC: 0 |
A batch of EU Regional tasks was released today. I haven't tried to download any, so can't confirm that they're actually available. If you have trouble downloading, please post. "We have met the enemy and he is us." -- Pogo Greetings from coastal Washington state, the scenic US Pacific Northwest. |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
There were 5,600 released. A few hours ago, there were about 3,600 left. A few minutes ago there was 2,495. Now there's 2,457. So they're being soaked up fast. Get one while you can. Backups: Here |
Send message Joined: 31 Dec 07 Posts: 1152 Credit: 22,363,583 RAC: 5,022 |
Boy those 5600 WU's went fast. If you Blinked you missed them. |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
According to the front page, there are 47,611 Active Hosts. Just 1 in 8 looking for work would do it. More available later, when they finish moving that 2 terabytes of data. Backups: Here |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
My tasks 'in progress' include 6 models (mainly HADAM3P) never received. Can these be reallocated somehow?Phantom models occur from time to time on this project and on others when the server load is high and the download process fails to complete properly. A result assignment record then appears but no corresponding entry is available in the BOINC Manager. As far as I know, it isn't possible to resend that work - at any rate, CPDN doesn't do it. So, no, the work cannot be resent to the original user. Allocation of the task to other users within the work unit will follow the normal BOINC rules. The project does from time to time re-issue work that hasn't been completed for one reason or another - but that's a manual task and would be done only for particularly interesting data regions or spin-ups (where completeness matters). The method of building parameter ensembles usually exploited by CPDN is inherently statistical: they neither require nor expect complete coverage. |
Send message Joined: 5 Aug 04 Posts: 127 Credit: 24,498,085 RAC: 21,454 |
Phantom models occur from time to time on this project and on others when the server load is high and the download process fails to complete properly. A result assignment record then appears but no corresponding entry is available in the BOINC Manager. As far as I know, it isn't possible to resend that work - at any rate, CPDN doesn't do it. Re-issuing of "lost" work was added as an option late 2005, so have been used by other BOINC-projects for over 5 years now. For a project it's enabled by including <resend_lost_results> in their normal project-config-file. The only other requirement is a minimum client-version, but this means v4.45, meaning 99.99% of active BOINC-users meets this part (and all CPDN-users since doesn't allow so old clients). Then enabled, re-issuing will send "lost" work to same computer next time computer asks for work (*), but within the limits of how much works asked for, and any other limits like max tasks to issue in one go, and of course doesn't re-issue work that can't fit within deadline or isn't needed any longer. So even if CPDN enables this option now, no-one will get 100+ old, "lost" models sent in one go. ;) The only downside with enabling re-issuing of "lost" work is that it adds extra load to database-server, so it's possible this will be too high load for CPDN to handle... (*): WCG runs by now very old server-code, so from this project you'll get re-issue even if don't ask for work, and will get all "lost" work at once. This won't be an issue for CPDN if enables <resend_lost_results>, since CPDN runs more resent code. |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
cpdn is different to other projects in many ways. One is that work units that fail on other projects can be re-sent because it's most likely been due to past-dead-line or hardware-failure issues. There's another reason here: the climate model's parameters may have sent the developing model into unrealistic conditions. An old favourite here, is "Negative pressure", where the atmospheric pressure goes negative. There are lots of others. So the data sets don't get re-sent, because they would most likely fail at the same point. And I don't think that BOINC has code that's sufficiently optioned to allow it to differentiate between the various failure modes. (Hardware/model parameter.) Backups: Here |
Send message Joined: 28 Mar 09 Posts: 126 Credit: 9,825,980 RAC: 0 |
The resend lost results only send the work that you never got. It compares a list of tasks your client says it has with what the project says its allocated to you. If you don't have one the project thinks was allocated then it will resend it. This is typically caused when there is a lot of network traffic and the scheduler reply to send you the work in the 1st place never gets through. If you got a work unit and its failed then it won't be resent, its reported as failed and the project will know you've had it and won't bother trying to give it back. As mentioned above it has a fair bit of database overhead, so may have been disabled simply for that reason, or maybe the project admin guys aren't aware of the option. |
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
As mentioned above it has a fair bit of database overhead, so may have been disabled simply for that reason, or maybe the project admin guys aren't aware of the option. They are aware of the option (or at least were before the recent staff changes) but it was, as you say, discounted because of the database overhead. Some server configuration changes are in progress (see the latest news post). Once the effect of those changes becomes apparent the decision not to resend lost results might be reviewed. "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
Send message Joined: 11 May 07 Posts: 36 Credit: 1,485,638 RAC: 0 |
Still no work - what is going on? My PC is connected and waiting during daylight hours and you'd think that some would come my way. Is the project coming to an end or were the moderators thinking that the End-Of-The-World prediction for the 21st May might be true... It would be useful to have an official position on when the 'No Work' dought will cease - Please Bill H |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Did you read my post for the 18th in this same thread? "Daylight hours" may have been why you missed them. And the current situation is described in the latest post in News and Announcements. Backups: Here |
Send message Joined: 6 Dec 08 Posts: 1 Credit: 369,809 RAC: 0 |
Between no new work and the stats not being updated it looks like this project is going by the way of the weather. Time to request this project to be removed from the DC-Vault. Proud Member of Team [H]ard|OCP & the [H]ard DC Commandos |
Send message Joined: 11 May 07 Posts: 36 Credit: 1,485,638 RAC: 0 |
Yes, I did read your post. I normally run 7am to 10pm but now have tried 24/7. Still no new work. While I think ClimatePrediction is a worthwhile use of spare PC resources (goodness knows we need better weather forcasting!) perhaps you have already collected all the data you need? Should I run alternative activities that maybe need the processor power more? Have we done enough and the research is nearing completion? It would be good to have an update... Bill H |
Send message Joined: 15 May 09 Posts: 4540 Credit: 19,039,635 RAC: 18,944 |
I would say that the problems in predicting exactly how the world's climate is going to change are a long way from being completely solved. The problem is that the project doesn't have enough money going into it to provide enough programmers and hardware. Compare the amount of money that has gone into climate prediction with that the nuclear industry gets for research. Of course climate prediction isn't tied in to the military so it will never get the same bucks. |
Send message Joined: 16 Jan 10 Posts: 1084 Credit: 7,827,799 RAC: 5,038 |
[Bill H wrote:] Have we done enough and the research is nearing completion? It would be good to have an update...That certainly isn't the case, if only because there are many different lines of research. The situation at the moment is rather like one of those embarrassing moments at a party when everyone stops talking just as you say "... I really loved that camel." There are more FAMOUS models on the way, but not quite yet. There are more HADCM3N models, but not just yet. I think there may be more HADAM3P models as well. There is a disk space problem (which is a bit odd since a new server was bought not too long ago) that is being sorted out; there is a relatively minor credit/message problem that is also being worked on but which really ought not stand in the way of new models (better to revert to old code than annoy users). So, there should be more models "soon". |
Send message Joined: 31 Dec 07 Posts: 1152 Credit: 22,363,583 RAC: 5,022 |
The problem is that the project doesn't have enough money going into it to provide enough programmers and hardware. Compare the amount of money that has gone into climate prediction with that the nuclear industry gets for research. Of course climate prediction isn't tied in to the military so it will never get the same bucks. Funny that you should mention the nuclear industry. I don’t know were you are live Dave, but, in the U.S. Federal budget proposals for F.Y. 2012 there is a proposal to essentially make a gift of 36 BILLION tax dollars to the U.S. nuclear industry under the guise of a “loan” program. The historic default rate on these kinds of loans is 50%. After what has just happened in Japan I am sure the default rate will go up. If only climate research got that kind of funding. |
©2024 cpdn.org