Questions and Answers :
Windows :
V8 very slow & wrong time
Message board moderation
Author | Message |
---|---|
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
hi everyone, i have version 8 running, i can download new wu's just fine, but ive noticed the wrong est.time remaining on most that i get. I am currently doing one that "did say" 7 hours when i first started the wu. that was almost 24 hours ago, it still says like 2 hours 30 minutes left. I have 4 other wu's that have est. time of over 19 hours each. all the wu's i have are v8, i have not been seeing any v7 wu's. Are all the v8 wu's that much larger than version 7's ? is it true seti@home is phasing out version 7 all together? (think i read that someplace here on the forums)
|
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
is it true seti@home is phasing out version 7 all together? You can see by yourself that no new SETI@home v7 WUs are created ("Results ready to send 0" and "Results out in the field 740,778" drop down every day): http://setiathome.berkeley.edu/sah_status.html  - ALF - "Find out what you don't do well ..... then don't do it!" :)  |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
hi everyone, i have version 8 running, i can download new wu's just fine, The times on tasks are estimates, will always be estimates. You have to run several - more than ten - for the estimate to become a more correct value. The project can not give out work with a correct estimate as they don't know what kind of computer you have. So they give out a general estimate, and BOINC will learn and adjust that number over the course of running these tasks. In the end you will notice that there are different tasks as well, all with different run times, but that when allowing BOINC to do its thing, in the end all of these different tasks will show their correct run times - still estimated - on your hardware. So it's just a patience thing. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
Hi BilBg, Thanks for pointing that out to me. Ive seen the page before but never really understood what a lot of it meant.
|
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
Hi Ageless. I understand the times are just estimates, although i always found version 7's estimates were more accurate. Made me curious when i seen one that says estimated 7 hrs and 23 hours later it still says estimated 2 hours remaining for the same wu. It does not seem right that a wu should take 2 days or more to complete. especially when version 7's normally completed in less than 10 hours. i dont know if its because of the size of the v8's wu's, if its normal for seti v8 to take that long. or if its a problem with my settings. (although settings were ok for v7)
|
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
Even for version 7 it took at least ten tasks without error and without being fully blanked, for the estimates to become near where the run time actually is. If the task you had was the problems with was the one you aborted, do know that your wingperson's computer took only 53 seconds to see this was a bad task, because of : Spike count: 24 It still does happen that tasks that should bounce out automatically don't do so on some computers and that their owner should interfere. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
hi Ageless, yes the ones i felt i was having issues with were the aborted ones. because they were taking much longer than they should have. Im sorry, i dont understand how to tell if its a bad task or not. you said do know that your wingperson's computer took only 53 seconds to see this was a bad task Sorry if this is a real noob question, but how do you tell its a bad task ?
|
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
You cannot tell if a task is a bad task, but the science application should normally when it finds too many occurrences of the things it's looking for: Spike count: All of these have certain thresholds. Go over and the task drops out early. They're not really bad tasks, that was a bit of an error from me, but they're tasks which have to be looked at by the project before they're resent to try again. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
Hi Ageless, thanks for the help. What about any size differences between seti v7 wu's and seti v8 wu's. Are there any differences ? or are both versions still generally the same.
|
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
Looking in the Seti directory, I see the new file size for version 8 work is 358KB, where it was 365-367 for version 7. Maybe something to do with the change in the Gaussian fitting routines. I'll ask around. Edit: I do know that eventually, when the Green Bank Telescope work is going to be run, that those tasks are in the order of 1 megabyte or more. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
hi Ageless, that seems odd, version 8 seems smaller than version 7. when i was running the version 7 wu's they worked very good on my system. then when i updated to version 8 is when all this seemed to start. ive only been at version 8 for about a week now and was only able to complete maybe 2 or 3 of the version 8 wu's i got so far, the rest were aborted. so i dont have a clue if its my system, version 8 or the wu's. i just know seti version 7 worked fine for me. edit: i wonder if it has to do with the fact im also using boinc tasks and not just boinc.
|
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
According to Richard Haselgrove: There's some extra configuration information in the <workunit_header>: An extra line to identify the task for v8 processing Several different splitter parameters, which can vary according to which telescope recorded the data - previously, only Arecibo recordings could be processed. The actual binary data after the XML header section is the same length. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
Thanks Ageless. I read that post but did not see anything to help me, i did see where it was referenced like you said though. But reading that i stumbled on info about lunatics. is lunatics working on a new enhanced client? I believe thats probably why im seeing my issues. (running boinc tasks without it) i did run lunatics with everything before while running the version 7 wu's. but this time i did a fresh install of boinc, boinc tasks and seti v8 without lunatics file. I may need to go back to basic boinc & seti 8 until lunatics is done. im just guessing.
|
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
thanks for the help guys, Not sure what happened, but after I installed boinc & boinc tasks again, (without lunatics file) everything went simple & easy. Downloaded 30+ new seti v8 wu's, all wu's show est. completion of 2 to 5 hours. (like they should)(not the 12 to 19+ hours i was previously seeing) all is good now thanks.
|
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14673 Credit: 200,643,578 RAC: 874 |
If by 'size' you meant 'estimated run time', then yes - when you first install Lunatics, you will see a (false) estimate for the first few dozen tasks. But if you watch the actual elapsed times, you should see that for tasks which normally run 2 - 5 hours (I'm guessing CPU tasks), the Lunatics applications are significantly quicker. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
Hi Richard, yes i run the 'cpu tasks', i was using the Lunatics application before, with seti 7, but when i started doing seti 8 tasks i did not realize i had to update the Lunatics application also. Since i learned the Lunatics application was not updated yet for seti 8 i thought i should do a fresh install of boinc, boinc tasks and seti 8. without the Lunatics application and wait for them to release an update.
|
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
...and have been running that for now until Lunatics application gets updated Which was yesterday: http://setiathome.berkeley.edu/forum_thread.php?id=78878, but before you quickly go update, set no new tasks first, run down the cache or abort everything, report everything and only then update to Lunatics. |
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
hi Ageless, thank you very much for that. very much appreciated. will do that the way you suggested. =)
|
xpozd Send message Joined: 26 Jan 15 Posts: 88 Credit: 280,183 RAC: 1 |
Ok now that ive tried to install the new Lunatics application im back to the previous issues again. 'cpu' tasks with est. completion times exceeding 19 to 20 hours each. (i received 12 tasks which says 9 days to complete) normally 12+ tasks would only show 3 days with seti v7 tasks when i click properties for current running tasks each show its an average of 30MB to 40MB for work unit size, does this seem right ? or am i once again getting a bunch of tasks that are 'bad'. but then its odd everything seemed ok before adding the Lunatics application. could Lunatics application be causing this?
|
rob smith Send message Joined: 7 Mar 03 Posts: 22441 Credit: 416,307,556 RAC: 380 |
Provided progress is being made at a sensible rate, that is one that will result in a run time comparable with that of your "non-Lunatics" installation then there is nothing to worry about. The estimate run time should be considered a "work of fiction", the one to watch is the actual elapsed time, and the progress. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
©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.