Returned wu piling up in pending queue


Advanced search

Message boards : Number crunching : Returned wu piling up in pending queue

Sort
Author Message
GeneM
Volunteer tester

Joined: Nov 28 06
Posts: 22
ID: 333
Credit: 7,402,034
RAC: 0
Message 1769 - Posted 14 Dec 2006 16:14:58 UTC

My pending credit queue is growing larger by the day. Since Dec. 9th it has grown larger everyday. It looks like about half to 2/3's of my returned work units just go into the pending credit queue. Is this normal?

Gene

Profile suguruhirahara
Forum moderator
Volunteer tester
Avatar

Joined: Sep 13 06
Posts: 282
ID: 15
Credit: 56,614
RAC: 0
Message 1770 - Posted 14 Dec 2006 16:30:12 UTC
Last modified: 14 Dec 2006 16:30:49 UTC

Hello,

For windows it's normal in general. Since the deadline of each workunit has been to set to 10 days, you'll have to wait for sending back of the other results in a workunit. For linux and mac because of lack of active hosts, you'll have to wait not only for return (uploaded to the server) of the other results in a workunit but also for sending to the hosts whose spec is relevant to your host, which is configured with homogenerous redundancy.

Hope this helps,
suguruhirahara
____________

I'm a volunteer participant; my views are not necessarily those of Docking@Home or its participating institutions.

Profile [B^S] Acmefrog
Volunteer tester
Avatar

Joined: Nov 14 06
Posts: 45
ID: 252
Credit: 1,604,407
RAC: 0
Message 1772 - Posted 14 Dec 2006 22:20:15 UTC

GeneM I know how you are feeling. I have had about a 1000 credits in waiting. Started to make me think that no one was crunching with XP.
____________

Profile David Ball
Forum moderator
Volunteer tester
Avatar

Joined: Sep 18 06
Posts: 274
ID: 115
Credit: 1,634,401
RAC: 0
Message 1774 - Posted 15 Dec 2006 1:48:29 UTC - in response to Message ID 1772 .

GeneM I know how you are feeling. I have had about a 1000 credits in waiting. Started to make me think that no one was crunching with XP.


I've seen similar results. Either people are gone for the holidays, have their computers working but not online, or have stopped running D@H.

I've seen one machine which was returning results within a day. It suddenly grabbed about 40 or 50 work units, returned two or three of them, and hasn't been heard from since. It might be they quit without aborting the work units or it might be that they just loaded the machine up to keep it busy, but not connected to the internet, while they went on holiday. I guess things are going to be strange this time of year.

It would be hard to tell with D@H, but I'd love to see what Windows Patch Tuesday does to an established project with thousands of machines. There are still people out here, like me, who live in small rural communities and are out of range for DSL, have no cable internet available, and are stuck on 56K dial-up. I spent most of a day downloading and re-booting my 2 WinXP machines.

Regards,

-- David
Profile suguruhirahara
Forum moderator
Volunteer tester
Avatar

Joined: Sep 13 06
Posts: 282
ID: 15
Credit: 56,614
RAC: 0
Message 1777 - Posted 15 Dec 2006 8:03:32 UTC - in response to Message ID 1772 .

GeneM I know how you are feeling. I have had about a 1000 credits in waiting. Started to make me think that no one was crunching with XP.

My pending credit list says more than 2800 credits are pended!:o
Profile David Ball
Forum moderator
Volunteer tester
Avatar

Joined: Sep 18 06
Posts: 274
ID: 115
Credit: 1,634,401
RAC: 0
Message 1778 - Posted 15 Dec 2006 12:48:00 UTC


Well, I have a little good news Gene. A lot of my pending work units on Linux have been crunched by me and usually by Conan. I just looked and they're now being handed mostly to you and someone called "Morgan the Gold". Since you'll complete the quorum, you should be getting credit on those as soon as you finish and the validator looks at them.

I'm up to 1658.58 on pending results. Hey, at least I can watch my point score rise without having just returned a result :-) I'm on dial-up so returning results for my crunching machines is a mostly manual process I do 3 or 4 times a day.

BTW, part of the problem may be lhcathome. They rarely hand out work units and when they do, it's with a short deadline so they push machines into EDF (earliest deadline first) and basically get priority service. They handed out some work units a few days ago.

Regards,

-- David

GeneM
Volunteer tester

Joined: Nov 28 06
Posts: 22
ID: 333
Credit: 7,402,034
RAC: 0
Message 1781 - Posted 15 Dec 2006 15:29:29 UTC

Hey,

Thanks guys for all the updates. I was really curious if something had gone wrong or not and I had just not heard about it. But it seems all is well however slow.

Gene

Jon
Volunteer tester

Joined: Nov 13 06
Posts: 1
ID: 209
Credit: 17,642
RAC: 0
Message 1782 - Posted 15 Dec 2006 16:40:56 UTC - in response to Message ID 1781 .

Hey,

Thanks guys for all the updates. I was really curious if something had gone wrong or not and I had just not heard about it. But it seems all is well however slow.

Gene



Thanks from me to I was wondering what was happen I thought I was the only one there for a while
Profile Andre Kerstens
Forum moderator
Project tester
Volunteer tester
Avatar

Joined: Sep 11 06
Posts: 749
ID: 1
Credit: 15,199
RAC: 0
Message 1785 - Posted 15 Dec 2006 22:18:21 UTC

Unfortunately it is true that there are lot of pending credits. More users with more machines would probably help a little bit, but we would really like to fix our problems with the apps first before letting more people attach. And David might be right about people having turned on their machines for the holidays: the university is awfully empty and quiet these days.... Hopefully the new year will bring new credit too :-)

Thanks
Andre
____________
D@H the greatest project in the world... a while from now!

Profile Conan
Volunteer tester
Avatar

Joined: Sep 13 06
Posts: 219
ID: 100
Credit: 4,256,493
RAC: 0
Message 1786 - Posted 16 Dec 2006 5:20:57 UTC

> While I have not crunched for Docking for a couple of days to concentrate on Rosetta, my Pending queue still has 74 WUs in it. This is down from over 100 so they are getting done. When it drops a bit more I will come back and start crunching again. It is mainly a Linux thing, Windows seems to only stay pending for a couple of days before being completed, Linux can take over 2 weeks.
____________

[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1790 - Posted 17 Dec 2006 6:49:20 UTC - in response to Message ID 1778 .
Last modified: 17 Dec 2006 6:51:54 UTC


Well, I have a little good news Gene. A lot of my pending work units on Linux have been crunched by me and usually by Conan. I just looked and they're now being handed mostly to you and someone called "Morgan the Gold". Since you'll complete the quorum, you should be getting credit on those as soon as you finish and the validator looks at them.

I'm up to 1658.58 on pending results. Hey, at least I can watch my point score rise without having just returned a result :-) I'm on dial-up so returning results for my crunching machines is a mostly manual process I do 3 or 4 times a day.

BTW, part of the problem may be lhcathome. They rarely hand out work units and when they do, it's with a short deadline so they push machines into EDF (earliest deadline first) and basically get priority service. They handed out some work units a few days ago.

Regards,

-- David


lol me ? I was wondering why my ears were ringing :)

Unfortunatly that machine got 'work bombed again'. I don't know why, it asks for 864 seconds of work, and gets 50 wu, (it's set at connect every .01 days). It has 30 left, all docking,takeing about 4 hours each.
____________

Memo
Forum moderator
Project developer
Project tester

Joined: Sep 13 06
Posts: 88
ID: 14
Credit: 1,666,392
RAC: 0
Message 1791 - Posted 17 Dec 2006 6:59:14 UTC - in response to Message ID 1786 .
Last modified: 17 Dec 2006 7:00:29 UTC

> While I have not crunched for Docking for a couple of days to concentrate on Rosetta, my Pending queue still has 74 WUs in it. This is down from over 100 so they are getting done. When it drops a bit more I will come back and start crunching again. It is mainly a Linux thing, Windows seems to only stay pending for a couple of days before being completed, Linux can take over 2 weeks.


Actually windows are suffering as well... but not much because the number of machines is higher. My pending has been coming down from almost 16K to 12K so that is a good indicator. I dont have many linux machines to help more but I will try to get some more.
[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1798 - Posted 17 Dec 2006 10:09:33 UTC

I don't know if this pc will work this time or not (and just chew up more wu's). If it does hopefully I can add a few more ;)
____________

Memo
Forum moderator
Project developer
Project tester

Joined: Sep 13 06
Posts: 88
ID: 14
Credit: 1,666,392
RAC: 0
Message 1803 - Posted 17 Dec 2006 15:22:39 UTC - in response to Message ID 1798 .

I don't know if this pc will work this time or not (and just chew up more wu's). If it does hopefully I can add a few more ;)


Did you change the stack size on this?
ulimit -s unlimited
[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1807 - Posted 18 Dec 2006 0:36:48 UTC - in response to Message ID 1803 .
Last modified: 18 Dec 2006 0:39:26 UTC

I don't know if this pc will work this time or not (and just chew up more wu's). If it does hopefully I can add a few more ;)


Did you change the stack size on this?
ulimit -s unlimited


doh, i try this now
the file i qsub was
#PBS -l nodes=1:blue:run

ulimit -s unlimited
cd $HOME
cd n
cd BOINC

tail -f /usr/spool/PBS/spool/$PBS_JOBID.OU > $PBS_O_WORKDIR/$PBS_JOBID.ou&
tail -f /usr/spool/PBS/spool/$PBS_JOBID.ER > $PBS_O_WORKDIR/$PBS_JOBID.er&

./boinc -exit_when_idle

echo "Donne!"


:(

____________

[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1811 - Posted 18 Dec 2006 7:12:25 UTC
Last modified: 18 Dec 2006 7:14:26 UTC

:) pending. Here's another

Profile suguruhirahara
Forum moderator
Volunteer tester
Avatar

Joined: Sep 13 06
Posts: 282
ID: 15
Credit: 56,614
RAC: 0
Message 1813 - Posted 18 Dec 2006 9:52:30 UTC - in response to Message ID 1811 .

:) pending. Here's another

The workunit seems to have no problem. Since the deadline is set to 10 days. If one/some of replica in a workunit won't be returned in the period the new copy will be distributed after that. So in that case you'd have to wait for 10 days additionally.

suguruhirahara
____________

I'm a volunteer participant; my views are not necessarily those of Docking@Home or its participating institutions.
[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1814 - Posted 18 Dec 2006 10:43:34 UTC
Last modified: 18 Dec 2006 10:46:38 UTC

lol I was just funnin' I added machines to help others have fewer pending and get new wu's instead :)

the first new pc finished of a quorum on 1 wu :)
____________

Profile David Ball
Forum moderator
Volunteer tester
Avatar

Joined: Sep 18 06
Posts: 274
ID: 115
Credit: 1,634,401
RAC: 0
Message 1815 - Posted 18 Dec 2006 11:44:45 UTC - in response to Message ID 1811 .

:) pending. Here's another


That work unit should finish about Thursday. GeneM, who started this thread *grin*, has the third machine. That machine has a long work queue so it takes just over 4 days on average to return the work unit but it's crunching away on them. It has returned over 5 work units in the last 24 hours so it's not on holiday :-)

-- David
[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1816 - Posted 18 Dec 2006 12:30:25 UTC
Last modified: 18 Dec 2006 12:39:16 UTC

i just noticed the wu's my 2100 is working on are marked over no reply :oops:
i thought i had 1 more day is late ok ?

[edit] it took the last few lol ,I'm about 4 days off,16 wu's due hours ago, and i aborted the ones i thought i wouldn't finish in time 4 days ago [/edit]
____________

Profile David Ball
Forum moderator
Volunteer tester
Avatar

Joined: Sep 18 06
Posts: 274
ID: 115
Credit: 1,634,401
RAC: 0
Message 1896 - Posted 29 Dec 2006 16:48:10 UTC

Now that the machines on holiday have timed out, my pending result queue has shrunk from about 1700 cobblestones to the 200 - 350 cobblestone range.

How are your pending queues doing?

-- David

GeneM
Volunteer tester

Joined: Nov 28 06
Posts: 22
ID: 333
Credit: 7,402,034
RAC: 0
Message 1897 - Posted 29 Dec 2006 17:40:26 UTC

My work queue has shrunk to 500 today. It was three times that yesterday though.

[B^S] Morgan the Gold
Volunteer tester
Avatar

Joined: Oct 2 06
Posts: 41
ID: 170
Credit: 138,735
RAC: 0
Message 1904 - Posted 30 Dec 2006 11:24:26 UTC

mine has dropped from 2000 to 1500, but 3 of my machines went on holidays 4 days ago.
____________

Message boards : Number crunching : Returned wu piling up in pending queue

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)#28 (2) {
      ["db_conn"]=>
      resource(102) 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=127"
    }
  }
  [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)#28 (2) {
      ["db_conn"]=>
      resource(102) of type (mysql link persistent)
      ["db_name"]=>
      string(7) "docking"
    }
    ["type"]=>
    string(2) "->"
    ["args"]=>
    array(3) {
      [0]=>
      object(BoincThread)#3 (16) {
        ["id"]=>
        string(3) "127"
        ["forum"]=>
        string(1) "2"
        ["owner"]=>
        string(3) "333"
        ["status"]=>
        string(1) "0"
        ["title"]=>
        string(38) "Returned wu piling up in pending queue"
        ["timestamp"]=>
        string(10) "1167477866"
        ["views"]=>
        string(4) "1498"
        ["replies"]=>
        string(2) "22"
        ["activity"]=>
        string(23) "4.2713980956043996e-126"
        ["sufferers"]=>
        string(1) "0"
        ["score"]=>
        string(1) "0"
        ["votes"]=>
        string(1) "0"
        ["create_time"]=>
        string(10) "1166112898"
        ["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) "127"
      ["forum"]=>
      string(1) "2"
      ["owner"]=>
      string(3) "333"
      ["status"]=>
      string(1) "0"
      ["title"]=>
      string(38) "Returned wu piling up in pending queue"
      ["timestamp"]=>
      string(10) "1167477866"
      ["views"]=>
      string(4) "1498"
      ["replies"]=>
      string(2) "22"
      ["activity"]=>
      string(23) "4.2713980956043996e-126"
      ["sufferers"]=>
      string(1) "0"
      ["score"]=>
      string(1) "0"
      ["votes"]=>
      string(1) "0"
      ["create_time"]=>
      string(10) "1166112898"
      ["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=127