some WU's take over 100% to complete


Advanced search

Message boards : Number crunching : some WU's take over 100% to complete

Sort
Author Message
BeemerBiker
Avatar

Joined: Aug 8 09
Posts: 3
ID: 16805
Credit: 692,240
RAC: 0
Message 5319 - Posted 17 Aug 2009 2:14:25 UTC

I joined this project a week ago and this is the first project I have seen where (occasionally) the work unit does not actually complete at 100%. Sometimes it continues to run, other times it is "waiting to run". Eventually it does complete though it may take 15 - 20 minutes beyond the "100%".

There are a number of projects that get stuck "uploading" and again, one simply waits it out and it eventually completes. I assume this is a similar problem.

Anyway, I was wondering if anyone else has noticed this.

Profile Scientific Frontline
Avatar

Joined: Mar 25 09
Posts: 42
ID: 8725
Credit: 788,015
RAC: 0
Message 5320 - Posted 17 Aug 2009 5:47:51 UTC - in response to Message ID 5319 .

I joined this project a week ago and this is the first project I have seen where (occasionally) the work unit does not actually complete at 100%. Sometimes it continues to run, other times it is "waiting to run". Eventually it does complete though it may take 15 - 20 minutes beyond the "100%".

There are a number of projects that get stuck "uploading" and again, one simply waits it out and it eventually completes. I assume this is a similar problem.

Anyway, I was wondering if anyone else has noticed this.

Can't say I have seen more then a minute or two, yet that is common in most projects. which OS are you having the problem with...
Although I really think it could be that you are from Texas!
OU#1
See you at the Red River. ~smiles~
Heidi-Ann
____________

Recognized by the Carnegie Institute of Science . Washington D.C.
Profile adrianxw
Volunteer tester
Avatar

Joined: Dec 30 06
Posts: 164
ID: 343
Credit: 1,669,741
RAC: 0
Message 5322 - Posted 17 Aug 2009 8:05:35 UTC
Last modified: 17 Aug 2009 8:06:40 UTC

If it is flagged as "Waiting to run" then BOINC has swapped tasks, not the application.

It is not uncommon for tasks here and at other projects to report 100% complete before they are ready to report. The crunching is done, but they, for example, are compressing the output files or similar, watch a CPDN wu complete for example - takes ages! For similar reasons, at some projects, a task starts and runs for several minutes whilst the "clock" stays 0.00%, it is unpacking and initialising structures.
____________
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

Profile Conan
Volunteer tester
Avatar

Joined: Sep 13 06
Posts: 219
ID: 100
Credit: 4,256,493
RAC: 0
Message 5324 - Posted 17 Aug 2009 12:30:15 UTC

As adrianxw says,
I see this on multiple projects very often.
It is just that Boinc has switched to another application as the time is up, it will finish quickly when it starts running again.

If you run just the one project you wont see this behaviour as Boinc wont be switching.
____________

BeemerBiker
Avatar

Joined: Aug 8 09
Posts: 3
ID: 16805
Credit: 692,240
RAC: 0
Message 5326 - Posted 17 Aug 2009 21:13:05 UTC - in response to Message ID 5324 .
Last modified: 17 Aug 2009 21:19:46 UTC

As adrianxw says,
I see this on multiple projects very often.
It is just that Boinc has switched to another application as the time is up, it will finish quickly when it starts running again.

If you run just the one project you wont see this behaviour as Boinc wont be switching.


If I understand correctly, then at the time the 100% is signaled, then boinc can deschedule the app and docking could sit for a long time before getting a time slice since it has not finished its housekeeping. IMHO, it seems to me that the 100% should be signaled after all housekeeping is done. I have not seen any other apps with this behavior. I do see a lot that get "stuck" uploading, mainly gpugrid with their humongus result file.

The first time I saw this (100.000 progress) I suspended all other tasks and docking went from "ready to run" to "running" and it was still stuck at 100.000 progress for at least 10 minutes. It was not stuck "uploading". I never did see it complete, mainly because I let the system reboot after installing the latest and greatest vista patches.

This seems to me to be an anomoly. Bringing up boincview, I see two full screens (about 100 or so and only 7 are docking) of "ready to report" and "100.000 progress". "Running" or "waiting to run" and "100.000 % progress" stick out like a sore thumb and are easy to spot. There are none right now.

Again, this has happened only 3 times. I think all three occurred on 2 windows systems.

Message boards : Number crunching : some WU's take over 100% to complete

Database Error
: The MySQL server is running with the --read-only option so it cannot execute this statement
array(3) {
  [0]=>
  array(7) {
    ["file"]=>
    string(47) "/boinc/projects/docking/html_v2/inc/db_conn.inc"
    ["line"]=>
    int(97)
    ["function"]=>
    string(8) "do_query"
    ["class"]=>
    string(6) "DbConn"
    ["object"]=>
    object(DbConn)#10 (2) {
      ["db_conn"]=>
      resource(72) of type (mysql link persistent)
      ["db_name"]=>
      string(7) "docking"
    }
    ["type"]=>
    string(2) "->"
    ["args"]=>
    array(1) {
      [0]=>
      &string(51) "update DBNAME.thread set views=views+1 where id=459"
    }
  }
  [1]=>
  array(7) {
    ["file"]=>
    string(48) "/boinc/projects/docking/html_v2/inc/forum_db.inc"
    ["line"]=>
    int(60)
    ["function"]=>
    string(6) "update"
    ["class"]=>
    string(6) "DbConn"
    ["object"]=>
    object(DbConn)#10 (2) {
      ["db_conn"]=>
      resource(72) of type (mysql link persistent)
      ["db_name"]=>
      string(7) "docking"
    }
    ["type"]=>
    string(2) "->"
    ["args"]=>
    array(3) {
      [0]=>
      object(BoincThread)#3 (16) {
        ["id"]=>
        string(3) "459"
        ["forum"]=>
        string(1) "2"
        ["owner"]=>
        string(5) "16805"
        ["status"]=>
        string(1) "0"
        ["title"]=>
        string(36) "some WU's take over 100% to complete"
        ["timestamp"]=>
        string(10) "1250543585"
        ["views"]=>
        string(3) "120"
        ["replies"]=>
        string(1) "4"
        ["activity"]=>
        string(19) "5.6341693292134e-85"
        ["sufferers"]=>
        string(1) "0"
        ["score"]=>
        string(1) "0"
        ["votes"]=>
        string(1) "0"
        ["create_time"]=>
        string(10) "1250475265"
        ["hidden"]=>
        string(1) "0"
        ["sticky"]=>
        string(1) "0"
        ["locked"]=>
        string(1) "0"
      }
      [1]=>
      &string(6) "thread"
      [2]=>
      &string(13) "views=views+1"
    }
  }
  [2]=>
  array(7) {
    ["file"]=>
    string(63) "/boinc/projects/docking/html_v2/user/community/forum/thread.php"
    ["line"]=>
    int(184)
    ["function"]=>
    string(6) "update"
    ["class"]=>
    string(11) "BoincThread"
    ["object"]=>
    object(BoincThread)#3 (16) {
      ["id"]=>
      string(3) "459"
      ["forum"]=>
      string(1) "2"
      ["owner"]=>
      string(5) "16805"
      ["status"]=>
      string(1) "0"
      ["title"]=>
      string(36) "some WU's take over 100% to complete"
      ["timestamp"]=>
      string(10) "1250543585"
      ["views"]=>
      string(3) "120"
      ["replies"]=>
      string(1) "4"
      ["activity"]=>
      string(19) "5.6341693292134e-85"
      ["sufferers"]=>
      string(1) "0"
      ["score"]=>
      string(1) "0"
      ["votes"]=>
      string(1) "0"
      ["create_time"]=>
      string(10) "1250475265"
      ["hidden"]=>
      string(1) "0"
      ["sticky"]=>
      string(1) "0"
      ["locked"]=>
      string(1) "0"
    }
    ["type"]=>
    string(2) "->"
    ["args"]=>
    array(1) {
      [0]=>
      &string(13) "views=views+1"
    }
  }
}
query: update docking.thread set views=views+1 where id=459