Message boards :
Number crunching :
BOINC 4.38 is out
Message board moderation
Author | Message |
---|---|
Raithmir Send message Joined: 3 Apr 99 Posts: 89 Credit: 385,065 RAC: 0 |
No problems with the repeated "May run out of work in 2.00 days; requesting more" error so far, but then I've only been running it 10 minutes! :) |
Scottatron Send message Joined: 15 Jul 03 Posts: 94 Credit: 220,389 RAC: 0 |
Have just installed 4.38 and have hit manual update numerous times, the client is not sending a request to the server. 10/05/2005 6:06:05 PM||request_reschedule_cpus: project op 10/05/2005 6:06:05 PM||schedule_cpus: must schedule 10/05/2005 6:06:05 PM||earliest deadline: 1116156049.000000 h0006A_1_62618_1 That's all it is doing, not sure if this is a feature or not, will wait and see, otherwise will role back to 4.32 again. |
Raithmir Send message Joined: 3 Apr 99 Posts: 89 Credit: 385,065 RAC: 0 |
No problems with that on mine, although I have noticed that Predictor is leaving the process in memory when pausing the process even though my preferences say not to. |
Scottatron Send message Joined: 15 Jul 03 Posts: 94 Credit: 220,389 RAC: 0 |
I think I might keep an eye on the forums over the next day or two to see how well it is received... |
Anthony Brixey Send message Joined: 24 Jun 00 Posts: 102 Credit: 1,757,916 RAC: 0 |
It looks like 4.38 has sorted out the ‘requesting more’ problem. Anthony |
Raithmir Send message Joined: 3 Apr 99 Posts: 89 Credit: 385,065 RAC: 0 |
Anyone else noticing it leaving threads in memory when not running? |
Captain Avatar Send message Joined: 17 May 99 Posts: 15133 Credit: 529,088 RAC: 0 |
<blockquote>Have just installed 4.38 and have hit manual update numerous times, the client is not sending a request to the server. 10/05/2005 6:06:05 PM||request_reschedule_cpus: project op 10/05/2005 6:06:05 PM||schedule_cpus: must schedule 10/05/2005 6:06:05 PM||earliest deadline: 1116156049.000000 h0006A_1_62618_1 That's all it is doing, not sure if this is a feature or not, will wait and see, otherwise will role back to 4.32 again.</blockquote> What projects are you running with 4.38? E@H has a no connect before your time instruction..... |
Yeti Send message Joined: 3 Jan 00 Posts: 19 Credit: 26,705,234 RAC: 0 |
<blockquote>10/05/2005 6:06:05 PM||earliest deadline: 1116156049.000000 h0006A_1_62618_1</blockquote> This is the important line. The client tells you, that he switched the internal-modus to "earliest deadline". This means, you are in danger to miss a deadline of one or more WUs; so the client does all he can to finish the WU(s) before the deadline is reached. As far as I know, the client will not download any new work until the modus "earliest deadline" is solved and switched back to "normal". This is one of the great benefits of the new scheduling-calculations. Supporting BOINC, a great concept ! |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
I noticed something different about the install of 4.38. I uninstalled 4.37, during that process, I was prompted to close my internet connection, then the uninstaller continued. When I installed 4.38 I was only prompted once by Mcafee firewall to allow access (this was for the manager). Then the window booted right up to show my work. On the previous two installs I got a "no connection" message box. This message occurred both (4.36 and 4.37) times directly after I granted access to the manager but before granting access to the client. On both of the previous installs, I got the empty tabs screen, until I shut the newly installed program down and restarted. 4.38 appears to be working so far. |
Scottatron Send message Joined: 15 Jul 03 Posts: 94 Credit: 220,389 RAC: 0 |
<blockquote><blockquote>Have just installed 4.38 and have hit manual update numerous times, the client is not sending a request to the server. 10/05/2005 6:06:05 PM||request_reschedule_cpus: project op 10/05/2005 6:06:05 PM||schedule_cpus: must schedule 10/05/2005 6:06:05 PM||earliest deadline: 1116156049.000000 h0006A_1_62618_1 That's all it is doing, not sure if this is a feature or not, will wait and see, otherwise will role back to 4.32 again.</blockquote> What projects are you running with 4.38? E@H has a no connect before your time instruction.....</blockquote> I am running S@H, PP@H and LHC. LHC = 500 resource share. Predictor = 400 resource share. SETI = 100 resourse share. |
Scottatron Send message Joined: 15 Jul 03 Posts: 94 Credit: 220,389 RAC: 0 |
<blockquote><blockquote>10/05/2005 6:06:05 PM||earliest deadline: 1116156049.000000 h0006A_1_62618_1</blockquote> This is the important line. The client tells you, that he switched the internal-modus to "earliest deadline". This means, you are in danger to miss a deadline of one or more WUs; so the client does all he can to finish the WU(s) before the deadline is reached. As far as I know, the client will not download any new work until the modus "earliest deadline" is solved and switched back to "normal". This is one of the great benefits of the new scheduling-calculations. </blockquote> However, what if I wanted to perform a manual update to update my credits/preferences etc? I just tried installing 4.38 again and performing/hitting update and all it does is this: 10/05/2005 10:19:26 PM||request_reschedule_cpus: project op 10/05/2005 10:19:26 PM||schedule_cpus: must schedule 10/05/2005 10:19:06 PM||earliest deadline: 1116273859.000000 v64D1D2MQonlyinjnoskew2b5off-30s12_14574.4231_1_sixvf_31975_2 Meaning: I have requested an update. The machine KNOWS it has to perform an update. BOINC then shows that an LHC WU has the earliest deadline. BUT! There is no update performed. Can someone else please try this, and post the results from their messages screen? Edit: I just performed another update at 10:25pm, and it then made a request to the appropriate server. Strange however..... |
Captain Avatar Send message Joined: 17 May 99 Posts: 15133 Credit: 529,088 RAC: 0 |
<blockquote>I noticed something different about the install of 4.38. I uninstalled 4.37, during that process, I was prompted to close my internet connection, then the uninstaller continued. When I installed 4.38 I was only prompted once by Mcafee firewall to allow access (this was for the manager). Then the window booted right up to show my work. On the previous two installs I got a "no connection" message box. This message occurred both (4.36 and 4.37) times directly after I granted access to the manager but before granting access to the client. On both of the previous installs, I got the empty tabs screen, until I shut the newly installed program down and restarted. 4.38 appears to be working so far.</blockquote> Don't forget access for BOINC.scr also. I actully got the same problem as always but this time McAfee I granted perms for the three and it worked ok, BOINC.exe Bonicmgr.exe Boinc.scr So far so good (untill the next time I reboot) |
Urban Send message Joined: 4 Jan 00 Posts: 26 Credit: 835,395 RAC: 0 |
<blockquote><blockquote>10/05/2005 6:06:05 PM||earliest deadline: 1116156049.000000 h0006A_1_62618_1</blockquote> This is the important line. The client tells you, that he switched the internal-modus to "earliest deadline". This means, you are in danger to miss a deadline of one or more WUs; so the client does all he can to finish the WU(s) before the deadline is reached. As far as I know, the client will not download any new work until the modus "earliest deadline" is solved and switched back to "normal". This is one of the great benefits of the new scheduling-calculations. </blockquote> Hi Yeti, "This is one of the great benefits of the new scheduling-calculations." I didn't agree with that because E@H's return time is half of Seti, but the calculation time is at least double, so E@H, will the most time be calculated! How is the "earliest deadline" mode calculated? Urban <a><img src="http://www.boincstats.com/stats/banner.php?cpid=3837f9fafc28ff2e9df5b13ae2f8aaf7"></a> A member of the <a><img src="http://www.boincstats.com/stats/teambanner.php?teamname=SETI.Germany"></a> team. |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
Update: Once an einstein WU was finished and ready to upload, I did finally get prompted to allow access to the client. This is definitely different behavior from the past. |
Yeti Send message Joined: 3 Jan 00 Posts: 19 Credit: 26,705,234 RAC: 0 |
<blockquote>"This is one of the great benefits of the new scheduling-calculations." I didn't agree with that because E@H's return time is half of Seti, but the calculation time is at least double, so E@H, will the most time be calculated! </blockquote>In the short-term-looking, you could be right, if your queue is filled up too much. But, in the long-term-looking, you will watch, that the client balances this out by not downloading WUs from the project, that has shorter deadlines.<blockquote>How is the "earliest deadline" mode calculated?</blockquote>Sorry, I didn't put this in my mind; I saw several post about this in the forums, please, take a look around Supporting BOINC, a great concept ! |
Scottatron Send message Joined: 15 Jul 03 Posts: 94 Credit: 220,389 RAC: 0 |
Just finished an LHC WU (Which did have the earliest deadline, no issues there). However, the WU has uploaded but not reported to the server, a manual update was requested (By me), and no sign of it, a full 7 minutes later.... |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
<blockquote>How is the "earliest deadline" mode calculated? </blockquote> See for a comprehensive explanation bottom of this post by JM7. |
mlcudd Send message Joined: 11 Apr 03 Posts: 782 Credit: 63,647 RAC: 0 |
However, the WU has uploaded but not reported to the server, a manual update was requested (By me), and no sign of it, a full 7 minutes later I am having the exact same problem. i have 21 Wu's reported but have not been able to "manually Update". It acts as though it is not even trying to contact the servers. I know there are HTTP settins and proxy setting under options, but I have never messed with these. Noithing is checked on any of the tabs. Could this be the problem? Respectfully, Rocky www.boincsynergy.com |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
I'm using dial up so I don't wait for automatic reporting. I use manual update. My Eine that finished this AM uploaded fine, and manual update reported the WU no problem. I'm not saying that you don't have an issue, only that it didn't happen to me yet. tony |
mlcudd Send message Joined: 11 Apr 03 Posts: 782 Credit: 63,647 RAC: 0 |
I'm using dial up so I don't wait for automatic reporting. I use manual update. My Eine that finished this AM uploaded fine, and manual update reported the WU no problem. I'm not saying that you don't have an issue, only that it didn't happen to me yet. Hey Tony, I am using dial up also. What does screen say when you manually update. Also, do you have anything pressed under the "tools" tab in the options. I always thought there was an option there that was just for "Automatic connection" but I could be wrong. This is what is say's when i try to manually update: 5/10/2005 9:01:26 AM||request_reschedule_cpus: project op 5/10/2005 9:01:26 AM||schedule_cpus: must schedule This happens everytime. I tried to change my preferences this morning also. that is how I noticed that it would not manually update. Thank You In Advance, Respectfully, Rocky www.boincsynergy.com |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.