Message boards :
Number crunching :
work availability
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 · Next
Author | Message |
---|---|
Send message Joined: 19 Feb 08 Posts: 14 Credit: 1,004,732 RAC: 773 |
Do I have to enable the test app to get workunits? Or will the production application be replaced with the new application? If yes - when? It worked fine for my Vista x64 machine but I am trying to d/l it to an XP 32 bit machine but it's not downloading the new app. I reset the project and deleted everything in the folder (same as the other machine but all I get is this. 11/30/2008 10:56:36 PM|Milkyway@home|Sending scheduler request: To fetch work. Requesting 80587 seconds of work, reporting 0 completed tasks 11/30/2008 10:56:41 PM|Milkyway@home|Scheduler request succeeded: got 0 new tasks 11/30/2008 10:56:41 PM|Milkyway@home|Message from server: No work sent Edit: Well this is strange, I deleted all the files in the milkyway project folder but I just downloaded 2 new WUs when there aren't supposed to be any, nor do I have the app file. I must have another project folder somewhere (the version is 6.2.19). 11/30/2008 11:08:09 PM|Milkyway@home|Sending scheduler request: To fetch work. Requesting 80673 seconds of work, reporting 0 completed tasks 11/30/2008 11:08:14 PM|Milkyway@home|Scheduler request succeeded: got 2 new tasks 11/30/2008 11:08:16 PM|Milkyway@home|Started download of parameters_generated_1227852900_66214 11/30/2008 11:08:16 PM|Milkyway@home|Started download of parameters_generated_1227852901_66221 11/30/2008 11:08:17 PM|Milkyway@home|Finished download of parameters_generated_1227852900_66214 11/30/2008 11:08:17 PM|Milkyway@home|Finished download of parameters_generated_1227852901_66221 11/30/2008 11:08:19 PM|Milkyway@home|Started download of stars_82.txt 11/30/2008 11:08:25 PM|Milkyway@home|Sending scheduler request: To fetch work. Requesting 103197 seconds of work, reporting 0 completed tasks 11/30/2008 11:08:30 PM|Milkyway@home|Scheduler request succeeded: got 0 new tasks Well for some reason I downloaded 2 of the old WUs despite not having an app nor an app_info.xml file. When they ran I got a computation error. 11/30/2008 11:18:34 PM|Milkyway@home|Starting gs_625_1227852900_66214_1 11/30/2008 11:18:34 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:35 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:35 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:36 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:37 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:38 PM|Milkyway@home|Computation for task gs_625_1227852900_66214_1 finished 11/30/2008 11:18:38 PM|Milkyway@home|Output file gs_625_1227852900_66214_1_0 for task gs_625_1227852900_66214_1 absent 11/30/2008 11:18:38 PM|Milkyway@home|Starting gs_629_1227852901_66221_1 11/30/2008 11:18:38 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:38 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:39 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:39 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:39 PM|Milkyway@home|[error] Process creation failed: The system cannot find the file specified. (0x2) 11/30/2008 11:18:40 PM|Milkyway@home|Computation for task gs_629_1227852901_66221_1 finished 11/30/2008 11:18:40 PM|Milkyway@home|Output file gs_629_1227852901_66221_1_0 for task gs_629_1227852901_66221_1 absent And some more. 11/30/2008 11:40:28 PM|Milkyway@home|Sending scheduler request: Requested by user. Requesting 116560 seconds of work, reporting 0 completed tasks 11/30/2008 11:40:33 PM|Milkyway@home|Scheduler request succeeded: got 8 new tasks 11/30/2008 11:40:36 PM|Milkyway@home|Started download of parameters_generated_1227855187_75889 11/30/2008 11:40:36 PM|Milkyway@home|Started download of parameters_generated_1227855187_75890 11/30/2008 11:40:38 PM|Milkyway@home|Finished download of parameters_generated_1227855187_75889 11/30/2008 11:40:38 PM|Milkyway@home|Started download of parameters_generated_1227855187_75891 11/30/2008 11:40:39 PM|Milkyway@home|Finished download of parameters_generated_1227855187_75890 11/30/2008 11:40:39 PM|Milkyway@home|Finished download of parameters_generated_1227855187_75891 11/30/2008 11:40:39 PM|Milkyway@home|Started download of parameters_generated_1227855189_75896 11/30/2008 11:40:39 PM|Milkyway@home|Started download of parameters_generated_1227855182_75818 11/30/2008 11:40:40 PM|Milkyway@home|Finished download of parameters_generated_1227855189_75896 11/30/2008 11:40:40 PM|Milkyway@home|Finished download of parameters_generated_1227855182_75818 11/30/2008 11:40:40 PM|Milkyway@home|Started download of parameters_generated_1227855189_75899 11/30/2008 11:40:40 PM|Milkyway@home|Started download of parameters_generated_1227855184_75844 11/30/2008 11:40:41 PM|Milkyway@home|Finished download of parameters_generated_1227855189_75899 11/30/2008 11:40:41 PM|Milkyway@home|Finished download of parameters_generated_1227855184_75844 11/30/2008 11:40:41 PM|Milkyway@home|Started download of parameters_generated_1227855184_75845 11/30/2008 11:40:42 PM|Milkyway@home|Finished download of parameters_generated_1227855184_75845 11/30/2008 11:40:44 PM|Milkyway@home|Sending scheduler request: To fetch work. Requesting 111925 seconds of work, reporting 0 completed tasks 11/30/2008 11:40:49 PM|Milkyway@home|Scheduler request succeeded: got 0 new tasks 11/30/2008 11:40:49 PM|Milkyway@home|Message from server: No work sent 11/30/2008 11:40:49 PM|Milkyway@home|Message from server: (reached per-CPU limit of 8 tasks) |
Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0 |
Sorry Travis but thats not the case. Just caught 2 as soon as they finished and they only gave about 10 credits each. Edit: 56859252 .... 24485 .... 1 Dec 2008 1:36:09 UTC .... 1 Dec 2008 7:19:57 UTC .... Over Success Done.... 2,125.09.... 4.51.... 9.97 A clear conscience is usually the sign of a bad memory |
Send message Joined: 7 Jun 08 Posts: 464 Credit: 56,639,936 RAC: 0 |
There's two types of 'stripe' searches going right now. One pays ~10, the other ~40. Here's Travis' post about it. Alinator |
Send message Joined: 4 Oct 08 Posts: 1734 Credit: 64,228,409 RAC: 0 |
There's two types of 'stripe' searches going right now. One pays ~10, the other ~40. Back to reality and normal credit levels |
Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0 |
There's two types of 'stripe' searches going right now. One pays ~10, the other ~40. Yep! While I liked the higher credits I think having everyone on a level field is better. And Travis can spend his time on THE app. instead of trying to keep everyone happy across multiple apps! A clear conscience is usually the sign of a bad memory |
Send message Joined: 7 Jun 08 Posts: 464 Credit: 56,639,936 RAC: 0 |
LOL... All good things must pass. :-) Still pays what SAH optimized does on my hosts though. Better than most. ;-) Alinator |
Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0 |
|
Send message Joined: 7 Jun 08 Posts: 464 Credit: 56,639,936 RAC: 0 |
Yeah, but it is nice to have the TDCF back down to something sane. It made scheduling for the CC tough when you had some where the FPOP estimates were in the ballpark and the others were off in left field. :-) I run minimum caches, so it didn't have much effect for me, but I bet some people were getting a little upset about that! ;-) Alinator |
Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0 |
|
Send message Joined: 4 Oct 08 Posts: 1734 Credit: 64,228,409 RAC: 0 |
I know it's less than 1 day since the official client switch, but the system seems to be much more stable with the two levels of WU - the short one (2.5 X Milksop's client time) and the long one (~4 X Milksop's client calculation time). Servers are not being hammered, the system is clear on all the Server Status parts and work now always (at least so far) seems to be available. Now we watch the RACs coming down from the stratosphere to normal levels over the next month. |
Send message Joined: 22 Mar 08 Posts: 90 Credit: 501,728 RAC: 0 |
|
Send message Joined: 19 Nov 07 Posts: 29 Credit: 3,353,124 RAC: 0 |
Edit: Well this is strange, I deleted all the files in the milkyway project folder but I just downloaded 2 new WUs when there aren't supposed to be any, nor do I have the app file. That's part of the problem: you manually deleted files. BOINC thinks the file is still there. But I think BOINC should notice files are missing and download them... maybe it doesn't do that for app files, only WU inputs? Please use "Reply" or "Quote" buttons on posts, instead of "reply to this thread". Keep the posts linked together ("X is a reply to Y"). |
Send message Joined: 4 Oct 08 Posts: 1734 Credit: 64,228,409 RAC: 0 |
Edit: Well this is strange, I deleted all the files in the milkyway project folder but I just downloaded 2 new WUs when there aren't supposed to be any, nor do I have the app file. Probably the best way of moving over, assuming all the WUs you had were crunched and reported, was to detach from the project and then reattach. The client, all support files and work would download automatically then start crunching. |
Send message Joined: 19 Feb 08 Posts: 14 Credit: 1,004,732 RAC: 773 |
Thanks, it worked - sorry for not thinking fo that myself. :) I presumed resetting the project would work but it didn't. The odd thing is that deleting all the files manually worked on another system, just not on this one. |
Send message Joined: 4 Oct 08 Posts: 1734 Credit: 64,228,409 RAC: 0 |
Glad to see things worked. I presume you downloaded plenty of work (to the 8 WUs per core max) OK? Since the change to the new official client the servers have kept pace nicely, and now nearly for 24 hours. Looks like the cruncher starvation might be a thing of the past. |
Send message Joined: 19 Feb 08 Posts: 14 Credit: 1,004,732 RAC: 773 |
No it has work from other projects and for now, BOINC is estimating the time to completion around 7.5 hours on a dual core and won't download more than a couple of WUs at a time. :p Once that ceases to be a problem my work download will have other issues. This system completes WUs in twice as long as my Q6700. If MW is run exclusively then the machine I'm typing on can get max 12 hours of work with the limit and the quad which is not permanently connected to the internet can only have 6 hours of work at one time. This machine is permanently connected so it's not a big problem. Getting all 8X4=32 WUs at once on my quad before I leave my apartment each morning will be necessary. I am working on getting all my machines some sort of permanent connection but without Cable or DSL my options are somewhat limited (I use a cellular 3G connection). It would be nice if the WU limits allowed for keeping a cache with 24 hours of work at one time (based on average completion times I guess). :p |
Send message Joined: 4 Oct 08 Posts: 1734 Credit: 64,228,409 RAC: 0 |
Presuming you used the Milksop fast client and compare your crunch times per WU back to that one. Then, as I run a QX6,700 Quad @ 3.0GHz - my time for the shorter WU was was 515 seconds using the Milksop client. Using the new official release client, which is doing a bit more science, my WU crunch time has risen to around 795 seconds (2.5 x as long). There is also a second WU which does considerable more science, and this takes ~ 55 minutes on the same PC (4 x the Milksop WU). This seems to be a little lower on some and slightly more on others. When I started crunching with the new official client the estimated time to completion was sky high compared to the real time. I cannot remember what it was, but crunching 8 and more brought the estimate quickly down to the real time, and my cache store lengthened as a result to the maximum of 8 WU oer core (including the one being crunched). |
Send message Joined: 19 Feb 08 Posts: 14 Credit: 1,004,732 RAC: 773 |
Just checking out my recent tasks on this machine I see I completed 13 tasks with the new client so far (or at leasts thats all I see) and my estimated time to completion with 5 tasks in the cache is still around 3 hours. My Q6700 (@2.66GHz) took about 44 minutes per new WU (roughly 4X as you said). On this machine I have not seen any of the WUs that are 2.5X as long. This dual core (T2300@1.6GHz) runs about 1.7 hours per WU, a bit more than twice as long as the Q6700. It still has a bit of catching up to do and while I suppose I could change the setting which makes the estimated times change faster, I'm wary of nasty surprises (e.g. Sztaki). |
Send message Joined: 12 Nov 07 Posts: 2425 Credit: 524,164 RAC: 0 |
Which of the new lengths of wu's is going to be the one used for time being? Doesn't expecting the unexpected make the unexpected the expected? If it makes sense, DON'T do it. |
Send message Joined: 4 Oct 08 Posts: 1734 Credit: 64,228,409 RAC: 0 |
All of them I think from my non-purged results per host and BOINC cache. But the 39.84 CS WU seems to be the one with the largest numbers reported |
©2024 Astroinformatics Group