Message boards : Number crunching : Daily quota
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
Profile Robert Pick

Send message
Joined: 1 Dec 05
Posts: 62
Credit: 11,398,274
RAC: 261
Message 22931 - Posted: 7 Sep 2011, 17:11:55 UTC

I just got this message and was wondering what it means.

9/7/2011 9:49:59 AM | LHC@home 1.0 | Sending scheduler request: Requested by user.
9/7/2011 9:49:59 AM | LHC@home 1.0 | Reporting 2 completed tasks, requesting new tasks for CPU and NVIDIA GPU
9/7/2011 9:50:01 AM | LHC@home 1.0 | Scheduler request completed: got 0 new tasks
9/7/2011 9:50:01 AM | LHC@home 1.0 | No work sent
9/7/2011 9:50:01 AM | LHC@home 1.0 | (reached daily quota of 60 tasks)
Can I only do 60 tasks a day?
ID: 22931 · Report as offensive     Reply Quote
Profile Krunchin-Keith [USA]
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 2 Sep 04
Posts: 209
Credit: 1,482,496
RAC: 0
Message 22932 - Posted: 7 Sep 2011, 17:22:39 UTC

It means just that.

Look under your account / computers for

Maximum daily WU quota per CPU 10/day

If a computer produces lots of errors the quotta drops so it does not waste so many tasks others can complete correctly. When it does them correct it can be raised. The project can set an upper limit, and to be fair to all those attached often does this so everybody can get thier fair share of work. Often too during startup until problems are all fixed this is set low, then when things are fine and they have abundant work they may raise it.
ID: 22932 · Report as offensive     Reply Quote
Profile Robert Pick

Send message
Joined: 1 Dec 05
Posts: 62
Credit: 11,398,274
RAC: 261
Message 22934 - Posted: 7 Sep 2011, 18:08:45 UTC - in response to Message 22932.  

Thanks Keith, I didn't know that! I did check to see if I was sending a lot of errors or invalids and saw not a one.
If I open up 2 more cores will it let me do 20 more? This is the first 8 core processor I've ever had and am a bit of a knothead!
Thanks again, Pick
ID: 22934 · Report as offensive     Reply Quote
Profile Krunchin-Keith [USA]
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 2 Sep 04
Posts: 209
Credit: 1,482,496
RAC: 0
Message 22989 - Posted: 10 Sep 2011, 16:10:47 UTC - in response to Message 22934.  

Thanks Keith, I didn't know that! I did check to see if I was sending a lot of errors or invalids and saw not a one.
If I open up 2 more cores will it let me do 20 more? This is the first 8 core processor I've ever had and am a bit of a knothead!
Thanks again, Pick

Sorry for late reply,

In theory yes. The limit is ## per day X number of active to boinc cpu cores. (not toal cpu cores in the machine)

However the limit has been changed a couple of times by the admin during testing here.

I'm not sure the current limit, i have various numbers across my machines of 2, 4, 10 and 20.

The limit also changes by the fact that if you constantly turn in bad work, it is reduced, when you turn in good work it is raised upto the limit set by the project, i'm not sure thou the increments it uses.

There is not a large amount of work at the moment, and it seems too many computers are attached, so it vanishes quickly.

I think if everyone is patient for the next week or two, all the problems will get sorted out, there is a planned server change (see the url thread), so after that is tested maybe we will get more work more frequently [cross fingers].
ID: 22989 · Report as offensive     Reply Quote
Profile Igor Zacharov
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 16 May 11
Posts: 79
Credit: 111,419
RAC: 0
Message 22992 - Posted: 10 Sep 2011, 19:14:07 UTC - in response to Message 22989.  

We have crunched quite a few jobs now. Thank you very much for your effort!

The rejection factor is small, but it may be worthwhile to understand the numerical cases behind it. We may need to take a pause in submitting new studies until this is understood. It may take us 3 or 4 days.

Another change - we should move machines for lhcathomeclassic on wednesday or thursday and also fix some small data base inconsistencies. If you have noticed problems, this is time to report it.

I know about some team-leader mismatches, some host-user mismatches, the missing statistics on the lhcathomeclassic/sixtrack page. Anything else?
skype id: igor-zacharov
ID: 22992 · Report as offensive     Reply Quote
Profile Krunchin-Keith [USA]
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 2 Sep 04
Posts: 209
Credit: 1,482,496
RAC: 0
Message 22993 - Posted: 10 Sep 2011, 20:57:38 UTC - in response to Message 22992.  

We have crunched quite a few jobs now. Thank you very much for your effort!

The rejection factor is small, but it may be worthwhile to understand the numerical cases behind it. We may need to take a pause in submitting new studies until this is understood. It may take us 3 or 4 days.

Another change - we should move machines for lhcathomeclassic on wednesday or thursday and also fix some small data base inconsistencies. If you have noticed problems, this is time to report it.

I know about some team-leader mismatches, some host-user mismatches, the missing statistics on the lhcathomeclassic/sixtrack page. Anything else?

I review the threads, all I can see is the front page has [Links....] and copyright that need to be repalced, and there is no server_status page which a lot of users have requested, here it says you need admin rights to view it, every other project has one freely viewable.
ID: 22993 · Report as offensive     Reply Quote
Profile Krunchin-Keith [USA]
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 2 Sep 04
Posts: 209
Credit: 1,482,496
RAC: 0
Message 22999 - Posted: 11 Sep 2011, 18:24:57 UTC

Something else to fix while you're fixing things:

David Anderson wrote:
There are (at least) two issues with 6.13.3:

1) For GPU apps, it passes an additional command-line argument that causes some apps to fail (usually by exiting immediately). This will be fixed in 6.13.4.

2) 6.13.3 handles file upload certificates (a mechanism that prevents DoS attacks on upload servers) differently than previous versions.
This change was necessary to make scheduler requests and replies readable by standard XML parsers. But it means that file uploads will fail to projects that a) use upload certificates, and b) aren't running current server code.

I sent email to boinc_projects describing this change, and recommending that projects disable upload certificates until they're able to upgrade their server code. Not all project admins read boinc_projects, so it may be necessary to contact them via their message boards or email, and I urge Alpha-testers to do so.

-- David


David Anderson, boinc_projects email list wrote:


"Upload certificates" are a mechanism that keeps bad guys from DoS'ing your upload servers (note: such an attack has never happened, as far as I know).

We're changing the format of upload certificates, and we're starting to test a version of the client for which old-format certificates won't work. Volunteers testers won't be able to upload completed jobs, and they may complain to you.

I suggest that all projects disable upload certificates. To do so, add the following to your config.xml file:

<dont_generate_upload_certificates/>
<ignore_upload_certificates/>

To resume using upload certificates, if you wish:

1) upgrade to the current server source code (from trunk)
2) wait for all jobs with old-format certificates to be dispatched
3) re-enable certificates by removing the above lines.

Let me know if any questions.

-- David
ID: 22999 · Report as offensive     Reply Quote
Toby Broom
Volunteer moderator

Send message
Joined: 27 Sep 08
Posts: 798
Credit: 644,672,204
RAC: 235,923
Message 23095 - Posted: 18 Sep 2011, 15:25:11 UTC

This is a great feature!!

Before the fast maching would get all the work and everyone else would miss out. Now everyone get a fair'er chance.
ID: 23095 · Report as offensive     Reply Quote
Profile Tom95134

Send message
Joined: 4 May 07
Posts: 250
Credit: 826,541
RAC: 0
Message 23097 - Posted: 18 Sep 2011, 16:49:18 UTC

You might want to consider making the "Daily Quota" self adjusting based on the number of Tasks waiting on the Server and the number of requests for Tasks that have occurred (recently). I noticed that there are now 15,000+ Tasks waiting today and this number has crept upwards over the past few days. I crunch Einstein, SETI, LHC 1.0, and Test4Theory (LHC 2.0 using an Oracle Virtual Box). I'd like to get a few more Tasks for LHC 1.0

I suspect that those Project members that had the big crunching farms have all gone to T4T or are being too limited by the quota to crunch waiting Tasks.

It looks like the quota is set to three Tasks per day.

Tom
ID: 23097 · Report as offensive     Reply Quote
Bill & Patsy
Avatar

Send message
Joined: 25 Apr 08
Posts: 11
Credit: 3,529,049
RAC: 0
Message 23098 - Posted: 18 Sep 2011, 17:00:44 UTC - in response to Message 23097.  

...
I suspect that those Project members that had the big crunching farms have all gone to T4T...
...
Tom

Not likely. T4T currently limits tasks to one per computer (not even one per cpu). Not much there for big crunching farms...

--Bill

ID: 23098 · Report as offensive     Reply Quote
Profile Igor Zacharov
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 16 May 11
Posts: 79
Credit: 111,419
RAC: 0
Message 23103 - Posted: 18 Sep 2011, 19:34:08 UTC - in response to Message 23098.  

On popular request I have increased the quotas on LHC@HOME 1.0 in config.xml:


< 20 WAS
---
> 40 NEW
74,75c74,75
< 5 WAS
< 3 WAS
---
> 10 NEW
> 10 NEW

Let's see what happens.
Any other suggestion of what can be improved?

skype id: igor-zacharov
ID: 23103 · Report as offensive     Reply Quote
Profile Robert Pick

Send message
Joined: 1 Dec 05
Posts: 62
Credit: 11,398,274
RAC: 261
Message 23122 - Posted: 18 Sep 2011, 23:01:50 UTC

Wow, massive change in the size of each WU! Will have to leave this unit run 24/7 to get these done before deadline. I will make changes to limit amount received. Still good to have things rolling in. Well not as panicky as I thought. Just checked the run times against progress. At first the time was constantly going up and progress was going nowhere. Seems to be turning around now. These are new and I have no experience with them and I can't find any other participants run times to compare. Oh well I'll do my best. Pick
ID: 23122 · Report as offensive     Reply Quote
Profile Tom95134

Send message
Joined: 4 May 07
Posts: 250
Credit: 826,541
RAC: 0
Message 23129 - Posted: 19 Sep 2011, 6:32:43 UTC - in response to Message 23103.  

On popular request I have increased the quotas on LHC@HOME 1.0 in config.xml:


< <daily_result_quota> 20 </daily_result_quota> WAS
---
> <daily_result_quota> 40 </daily_result_quota> NEW
74,75c74,75
< <max_wus_to_send> 5 </max_wus_to_send> WAS
< <max_wus_in_progress> 3 </max_wus_in_progress> WAS
---
> <max_wus_to_send> 10 </max_wus_to_send> NEW
> <max_wus_in_progress> 10 </max_wus_in_progress> NEW

Let's see what happens.
Any other suggestion of what can be improved?

Bless you! :)
ID: 23129 · Report as offensive     Reply Quote
Profile Robert Pick

Send message
Joined: 1 Dec 05
Posts: 62
Credit: 11,398,274
RAC: 261
Message 23178 - Posted: 21 Sep 2011, 18:14:46 UTC

WoW, it was running HSW for a while till I hit the Quota limit again. I ran 320 in about 10 Min. Most for .02 credits. Now I have to wait for ever to run some more. Hope I don't miss too many that have some real credit!!! Pick
ID: 23178 · Report as offensive     Reply Quote
[AF>FAH-Addict.net]toTOW

Send message
Joined: 9 Oct 10
Posts: 77
Credit: 3,671,357
RAC: 0
Message 23180 - Posted: 21 Sep 2011, 18:33:48 UTC - in response to Message 23178.  

WoW, it was running HSW for a while till I hit the Quota limit again. I ran 320 in about 10 Min. Most for .02 credits. Now I have to wait for ever to run some more. Hope I don't miss too many that have some real credit!!! Pick


I think something is wrong in the current batch : the WUs are estimated to run 8 hours, but terminate only within seconds ...
ID: 23180 · Report as offensive     Reply Quote
metalius
Avatar

Send message
Joined: 3 Oct 06
Posts: 101
Credit: 8,985,206
RAC: 43
Message 23181 - Posted: 21 Sep 2011, 19:09:54 UTC - in response to Message 23180.  

I think something is wrong in the current batch the WUs are estimated to run 8 hours, but terminate only within seconds ...

If current input parameters of the simulation are giving unstable trajectories of the particle beam (the beam will hit something inside of LHC), there is no reason to continue this simulation - the task is finished. So this batch may be Ok with high probability.
ID: 23181 · Report as offensive     Reply Quote
Profile Nawiedzony
Avatar

Send message
Joined: 4 Jul 07
Posts: 17
Credit: 35,310
RAC: 0
Message 23182 - Posted: 21 Sep 2011, 19:21:51 UTC - in response to Message 23181.  

Yes I know about that ("beam hit the wall"), but it still sucks, when all yours daily quota is lost for 5s WUs...
ID: 23182 · Report as offensive     Reply Quote
metalius
Avatar

Send message
Joined: 3 Oct 06
Posts: 101
Credit: 8,985,206
RAC: 43
Message 23184 - Posted: 21 Sep 2011, 20:26:09 UTC - in response to Message 23182.  

Nawiedzony,
You are right, but this is our job here - our hosts must do simulations until correct parameters for shooting will be found. :-)
Possible solution…
Let's ask the project team to increase daily quota!
We have already very important and useful for the project limitation - limited amount of tasks in progress. By this limitation there is no possibility for some undecided participants to download hundreds of tasks and, for example, uninstall BOINC after that. IMHO, daily quota is an extra-limitation now and may be set to formal (big) value.

ID: 23184 · Report as offensive     Reply Quote
Eric Mcintosh
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 12 Jul 11
Posts: 857
Credit: 1,619,050
RAC: 0
Message 23185 - Posted: 21 Sep 2011, 21:32:15 UTC

terribly sorry; MEA CULPA. Put it down to old age,
stress, overwork, or whatever. I put in a whole batch of
rubbish jobs. I should have tested them first. I suspect a real
physics problem with a bad setup of TUNES perhaps.
promise to do better, and there is more genuine 10**6
turn jobs on their way. Sorry abot the 60 limit.
On the positive this is an excellent stress test. I used
to measure the overhead of batch systems but seeing how
long it took to do absolutely nothing.

We will also have to have a timeout soon to check the
results with ifort as against our old lahey lf95. Apart
from testing the system this is a major objective of the
current runs.

Eric.


ID: 23185 · Report as offensive     Reply Quote
Eric Mcintosh
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 12 Jul 11
Posts: 857
Credit: 1,619,050
RAC: 0
Message 23186 - Posted: 21 Sep 2011, 21:57:15 UTC

Well maybe I was too quick there. I have just run a couple
of cases and they are OK. Except that the particles don't
even complete one turn! I'll have to consult colleagues on
this. I should also wait for all results.

It is sad that if you get a "good" million turn job you'll
get a nice credit, but if you get a bad one you are immediately
available to get another one and can soon use up your quota of 60.

Have to think about this one. Eric.


ID: 23186 · Report as offensive     Reply Quote
1 · 2 · 3 · Next

Message boards : Number crunching : Daily quota


©2024 CERN