log in

Work almost done?


Advanced search

Message boards : Number crunching : Work almost done?

1 · 2 · Next
Author Message
Profile Miska
Send message
Joined: 30 Dec 05
Posts: 1
Credit: 13,154
RAC: 0
Message 12195 - Posted: 20 Jan 2006, 18:14:03 UTC

are we running out of work soon? As you can see works ready to crunch is decreasing quite rapidly
(yesterday over 300000 and now 177701).
____________

senatoralex85
Send message
Joined: 17 Sep 05
Posts: 60
Credit: 4,221
RAC: 0
Message 12196 - Posted: 20 Jan 2006, 18:58:52 UTC - in response to Message 12195.
Last modified: 20 Jan 2006, 18:59:08 UTC

Good Observation! I do not know. Maybe?

are we running out of work soon? As you can see works ready to crunch is decreasing quite rapidly
(yesterday over 300000 and now 177701).


____________

jsprague78
Send message
Joined: 10 Dec 05
Posts: 12
Credit: 8,932
RAC: 0
Message 12199 - Posted: 21 Jan 2006, 0:57:06 UTC - in response to Message 12195.

are we running out of work soon? As you can see works ready to crunch is decreasing quite rapidly
(yesterday over 300000 and now 177701).



I've noticed that as well, now down to about 158,000. Even if it goes dormant again for a few weeks it's been a pleasure to be able to participate in the projects, just wish I could participate in the LHC Alpha too.

jsprague78
Send message
Joined: 10 Dec 05
Posts: 12
Credit: 8,932
RAC: 0
Message 12200 - Posted: 21 Jan 2006, 0:57:06 UTC - in response to Message 12195.
Last modified: 21 Jan 2006, 0:57:49 UTC

Ignore, finger twitched when I went to click on Post Reply.

Desti
Send message
Joined: 16 Jul 05
Posts: 80
Credit: 694,086
RAC: 372
Message 12201 - Posted: 21 Jan 2006, 1:17:09 UTC

I'm sure > 50 % of that units are not done yet, because they were eaten by monster caches. 0_o
____________
Linux Users Everywhere @ BOINC
[url=http://lhcathome.cern.ch/team_display.php?teamid=717]

Profile Paul D. Buck
Send message
Joined: 2 Sep 04
Posts: 545
Credit: 148,912
RAC: 0
Message 12213 - Posted: 21 Jan 2006, 13:17:04 UTC

The historical pattern is to have "waves" of work, where a couple hundred thousand work units will be created and mostly issued, then another wave, if any, is added. After all the main waves have been issued and mostly returned, the "clean up" runs for a little bit where smaller batches are scavenged of work that is still incomplete (no quorum).
____________

John McLeod VII
Avatar
Send message
Joined: 2 Sep 04
Posts: 165
Credit: 143,002
RAC: 2
Message 12238 - Posted: 21 Jan 2006, 23:26:51 UTC

Then there is a downtime of a few days / weeks, and more is issued. I suspect that this pattern will continue until the construction is complete. At which point, I hope that we will be given some of the particle tracks to follow.
____________


BOINC WIKI

borandi
Send message
Joined: 9 Nov 05
Posts: 45
Credit: 20,379
RAC: 0
Message 12246 - Posted: 22 Jan 2006, 1:36:14 UTC - in response to Message 12213.

The historical pattern is to have "waves" of work, where a couple hundred thousand work units will be created and mostly issued, then another wave, if any, is added. After all the main waves have been issued and mostly returned, the "clean up" runs for a little bit where smaller batches are scavenged of work that is still incomplete (no quorum).


It leaves those who have done work waiting some weeks to pick up those last credits :D
____________
BOINC@Hull . xfire . mySpace

Feisal
Send message
Joined: 16 Jul 05
Posts: 22
Credit: 4,438
RAC: 0
Message 12259 - Posted: 22 Jan 2006, 14:00:57 UTC - in response to Message 12238.

Then there is a downtime of a few days / weeks, and more is issued. I suspect that this pattern will continue until the construction is complete. At which point, I hope that we will be given some of the particle tracks to follow.


How the hell would they be able to convert 1.2 gigabytes/s of work into to several kilobytes for us to be able to work. The work unti would pile up to much. Thats why they are creating the Grid. It can handle that kinda crazy power out put.
____________

Profile Paul D. Buck
Send message
Joined: 2 Sep 04
Posts: 545
Credit: 148,912
RAC: 0
Message 12264 - Posted: 22 Jan 2006, 14:59:11 UTC - in response to Message 12259.

[How the hell would they be able to convert 1.2 gigabytes/s of work into to several kilobytes for us to be able to work. The work unti would pile up to much. Thats why they are creating the Grid. It can handle that kinda crazy power out put.

Which is exactly what BOINC is ... a grid computing solution.

I have the room on disk, I have 9 reasonably fast computers, the cost is right, the only question is if this is the way they want to do the work... I think that if the DID start, and allowed more sign-ups ... we would continue to grow ... many don't do LHC@Home because the work is so intermittent ...
____________

Profile Rebirther
Avatar
Send message
Joined: 18 Sep 04
Posts: 18
Credit: 26,904
RAC: 0
Message 12270 - Posted: 22 Jan 2006, 17:02:26 UTC

And the deadline is too short...
____________

Profile Paul D. Buck
Send message
Joined: 2 Sep 04
Posts: 545
Credit: 148,912
RAC: 0
Message 12286 - Posted: 23 Jan 2006, 3:20:11 UTC - in response to Message 12270.

And the deadline is too short...

It is what the project needs/desires ... not what we would like ...
____________

senatoralex85
Send message
Joined: 17 Sep 05
Posts: 60
Credit: 4,221
RAC: 0
Message 12287 - Posted: 23 Jan 2006, 3:23:40 UTC - in response to Message 12264.

[How the hell would they be able to convert 1.2 gigabytes/s of work into to several kilobytes for us to be able to work. The work unti would pile up to much. Thats why they are creating the Grid. It can handle that kinda crazy power out put.

Which is exactly what BOINC is ... a grid computing solution.

I have the room on disk, I have 9 reasonably fast computers, the cost is right, the only question is if this is the way they want to do the work... I think that if the DID start, and allowed more sign-ups ... we would continue to grow ... many don't do LHC@Home because the work is so intermittent ...


-----------------------------------------------

I looked in the top 100 stats for boinc. Out of the top 100 crunchers crunching for more than one project, less than 5 were donating time to LHC. Of those users, most were donating less than 10 percent of their resources. WOW! Interesting to me!

____________

Profile Paul D. Buck
Send message
Joined: 2 Sep 04
Posts: 545
Credit: 148,912
RAC: 0
Message 12288 - Posted: 23 Jan 2006, 3:36:06 UTC

Yeah, and I probably fell off the list ...

I seemed to have dropped 12 in world position in just the last week ...

And I am ONLY giving LHC 21.61% ... it would be more if I was not trying to get EAH over SETI ... if I had REAL confidence that LHC would become a more "steady" project I would probably have given more here ... but, I don't have hopes I could realistically get enough work to get LHC over SETI before they run out of work ...

And, so, i do other projects a lot too ...
____________

Profile Yeti
Avatar
Send message
Joined: 2 Sep 04
Posts: 163
Credit: 10,496,269
RAC: 59,583
Message 12301 - Posted: 23 Jan 2006, 11:29:50 UTC

HM, I'm participating in nearly all BOINC-Projects.

When LHC sends work, I switch all other projects to "no new work". So, if LHC has work, my boxes concentrate on LHC :-)

____________


Supporting BOINC, a great concept !

Profile Keck_Komputers
Send message
Joined: 1 Sep 04
Posts: 275
Credit: 1,049,897
RAC: 967
Message 12314 - Posted: 23 Jan 2006, 16:25:28 UTC - in response to Message 12301.

HM, I'm participating in nearly all BOINC-Projects.

When LHC sends work, I switch all other projects to "no new work". So, if LHC has work, my boxes concentrate on LHC :-)

I just set it and forget it. I usually build up LTD in the outages and therefore normally have work when it is available.
____________
BOINC WIKI

BOINCing since 2002/12/8

Profile Ananas
Send message
Joined: 17 Jul 05
Posts: 102
Credit: 542,016
RAC: 0
Message 12326 - Posted: 23 Jan 2006, 19:29:36 UTC
Last modified: 23 Jan 2006, 19:30:58 UTC

If LTD is the same as <debt>, then at least 4.19 does not increase it while no work is available. I'm attached to Orbit for quite some time now and never had anything but 0.000 in the <debt> entry. Same for SZTAKI while it didn't give me work, it went to 0.000 and stayed there.

borandi
Send message
Joined: 9 Nov 05
Posts: 45
Credit: 20,379
RAC: 0
Message 12328 - Posted: 23 Jan 2006, 21:03:23 UTC - in response to Message 12301.

HM, I'm participating in nearly all BOINC-Projects.

When LHC sends work, I switch all other projects to "no new work". So, if LHC has work, my boxes concentrate on LHC :-)


Me too =D Mind you, there is now less than 15k WU left :'(
____________
BOINC@Hull . xfire . mySpace

John McLeod VII
Avatar
Send message
Joined: 2 Sep 04
Posts: 165
Credit: 143,002
RAC: 2
Message 12332 - Posted: 23 Jan 2006, 22:08:48 UTC - in response to Message 12326.

If LTD is the same as <debt>, then at least 4.19 does not increase it while no work is available. I'm attached to Orbit for quite some time now and never had anything but 0.000 in the <debt> entry. Same for SZTAKI while it didn't give me work, it went to 0.000 and stayed there.

4.19 was before the split between Short Term Debt and Long Term Debt. What was used in 4.19 is Short Term Debt and was called Debt.
____________


BOINC WIKI

River~~
Send message
Joined: 13 Jul 05
Posts: 456
Credit: 68,968
RAC: 0
Message 12348 - Posted: 24 Jan 2006, 6:15:31 UTC - in response to Message 12326.

If LTD is the same as <debt>, then at least 4.19 does not increase it while no work is available. I'm attached to Orbit for quite some time now and never had anything but 0.000 in the <debt> entry. Same for SZTAKI while it didn't give me work, it went to 0.000 and stayed there.


With the latest clients LTD behaves differently depending if +ve or -ve.

Large -ve debt means a project gets temporarily barred from seeking work, in effect it has had more than its fair share in the past. This will reduce to zero as the other projects work through their +ve debt, irrespective of whether the barred project has work.

If you think about this it is obvious - if the client is not asking project X for work it doesn't know there is none.

When the debt is close to zero or greater than zero, the project starts asking for work, and LTD then does not increase further whether there is work or not. LTD only increases past zero when *other* projects start hogging the cpu in EDF mode.

I guess what John Keck is seeing is that his computers download enough LHC units to go into EDF and build up a negative debt while there is work, and that the gaps in LHC work allow debt to "rise" back to zero.

Or just maybe he is using an old client - there was at least one release where the debt went on rising past zero when there was no work - a behaviour I preferred but apparently others shouted it down so it got changed (but that is a different argument)

____________

1 · 2 · Next

Message boards : Number crunching : Work almost done?