The At-De-At-tach ... concept!

Message boards : Number crunching : The At-De-At-tach ... concept!

To post messages, you must log in.

AuthorMessage
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10274 - Posted: 31 Jan 2006, 19:01:44 UTC

Hi ...

I keep on receiving the ...
1/31/2006 9:44:27 PM|rosetta@home|Note: not requesting new work or reporting results
...everytime I make an "update"...even when my work cache is completely empty and no matter how many times I "reset" the project?

Network usage
Connect to network about every
(determines size of work cache; maximum 10 days) 2 days

Computer summary
CPU type GenuineIntel Intel(R) Pentium(R) 4 CPU 3.00GHz
Number of CPUs 2
Operating System Microsoft Windows XP
Professional Edition, Service Pack 2, 05.01.2600.00)

Memory 501.96 MB
Cache 976.56 KB
Swap space 1228.68 MB
Total disk space 39.06 GB
Free Disk Space 31.5 GB
Use no more than 100 GB disk space
Leave at least 0.1 GB disk space free
Use no more than 50% of total disk space
Use no more than 75% of total virtual memory

Measured floating point speed 1920.73 million ops/sec
Measured integer speed 2909.75 million ops/sec
Average turnaround time 2.47 days
Maximum daily WU quota per CPU 100/day
Last time contacted server 31 Jan 2006 18:44:25 UTC
Average CPU efficiency 0.969474
Result duration correction factor 1.067268

Thanking you in advance

O&O
ID: 10274 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Moderator9
Volunteer moderator

Send message
Joined: 22 Jan 06
Posts: 1014
Credit: 0
RAC: 0
Message 10275 - Posted: 31 Jan 2006, 19:08:23 UTC - in response to Message 10274.  
Last modified: 31 Jan 2006, 19:08:58 UTC

O&O wrote:
Hi ...

I keep on receiving the ...
1/31/2006 9:44:27 PM|rosetta@home|Note: not requesting new work or reporting results
...everytime I make an "update"...even when my work cache is completely empty and no matter how many times I "reset" the project? ...



Any time you initiate an update through manual intervention such as clicking "Update" on the projects tab of the BOINC manager, you will get that message unless you actually have results to upload.

All it means is that the update contact with the server was not for the purpose of automatically obtaining new work or uploading and reporting any results.

If you are running other project on that machine, it may be that there is so much work present that the machine would not have time to load new work and still meet the deadlines for new R@H work. That would prevent the download until such time as the system could free up some time to devote to R@H.

Moderator9
ROSETTA@home FAQ
Moderator Contact
ID: 10275 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10278 - Posted: 31 Jan 2006, 19:36:36 UTC - in response to Message 10275.  
Last modified: 31 Jan 2006, 19:50:56 UTC

Hi ...

Any time you initiate an update through manual intervention such as clicking "Update" on the projects tab of the BOINC manager, you will get that message unless you actually have results to upload.

All it means is that the update contact with the server was not for the purpose of automatically obtaining new work or uploading and reporting any results.


Then, ...when will the client be able to automatically report/upload results and obtain new ones when there is NOW no "work" to work on?

If you are running other project on that machine, it may be that there is so much work present that the machine would not have time to load new work and still meet the deadlines for new R@H work. That would prevent the download until such time as the system could free up some time to devote to R@H.


I said "even when my work cache is completely empty"... and I meant "empty" from WUs of whatsoever!

What's your definition for the right "time"?

O&O (I was only able to read your reply when I hit the [Reply to this post],... I'm color blind you know)
ID: 10278 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10283 - Posted: 31 Jan 2006, 20:10:17 UTC - in response to Message 10278.  
Last modified: 31 Jan 2006, 20:41:18 UTC

(To Make things easy)

I run ...
R@H with 400% Rsource Share and 2 days Work Cache
Pirates with 200% RS and 2 days WC
And PrimeGrid with 400% RS and 2 days WC

... and I "reset" the three project one after another in the same order above ... then I hit the "Retry communications" through the BOINC client/ Command tab ...

Knowing that Pirates rarely have WU to send, guess which project received 345600 seconds of new work with 2/14/2006 as deadline and which one reported ... "Note: not requesting new work or reporting results"?

O&O
ID: 10283 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
John McLeod VII
Avatar

Send message
Joined: 17 Sep 05
Posts: 108
Credit: 195,137
RAC: 0
Message 10301 - Posted: 1 Feb 2006, 3:01:25 UTC

Do you have work from any project? If so, how much? What are the deadlines? What is the long term debt of the projects (from the client_state.xml, BOINCView, or BOINCDebtView)?


BOINC WIKI
ID: 10301 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10302 - Posted: 1 Feb 2006, 3:04:13 UTC - in response to Message 10301.  
Last modified: 1 Feb 2006, 3:15:11 UTC

Do you have work from any project? If so, how much? What are the deadlines? What is the long term debt of the projects (from the client_state.xml, BOINCView, or BOINCDebtView)?


Now I have 315900 (3:14:59 x 27) seconds of PrimeGrid with 2/14/2006 as deadline.

And BOINC DV reports the following LTD ...

PRJ: ROSETTA@HO STD: 0.000000 LTD: -21921.909536 RSRC: 400 -------------------------------
PRJ: PIRATES@HO STD: 0.000000 LTD: 105015.606196 RSRC: 200 -------------------------------
PRJ: PRIMEGRID STD: 0.000000 LTD: -83093.696659 RSRC: 400 -------------------------------

O&O
ID: 10302 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10516 - Posted: 6 Feb 2006, 23:26:37 UTC
Last modified: 6 Feb 2006, 23:38:40 UTC

an update:

Living with the "note" everytime I'm out of R@H WUs, I finally managed to have me some... and the secret is to Detach from the the project and then re-attach to it again (in a matter of seconds) without changing my previous prefrences without quitting any other project.

And guess what?...17 WUs are currently downloading together with Rosetta 4.81 application.

The question is, if the client finds that my computer has a room for this amount of R@H works then why it gives the "Note: not requesting new work or reporting results" when my computer runs out of it?

O&O (Adopting the new Rosetta life concept...the At-De-At-tach)
ID: 10516 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Paul D. Buck

Send message
Joined: 17 Sep 05
Posts: 815
Credit: 1,812,737
RAC: 0
Message 10521 - Posted: 7 Feb 2006, 1:26:16 UTC

Becasuse, based on your preferences and your resource shares etc. The project should not send you work. When you detach and reattach, 2 seconds or 10 years, these demand/"debt" numbers are set to zero.

So, what you are doing is forcing BOINC to ignore your wish to spend 20% of your time in RAH (for example)...

BOINC balances work OVER TIME, so there will be times when you will have no work from some projects ...
ID: 10521 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10525 - Posted: 7 Feb 2006, 3:46:47 UTC - in response to Message 10521.  

Becasuse, based on your preferences and your resource shares etc. The project should not send you work.


I see

...And based on your knowledge and respect to the science which BOINC resemples, what modification should I do to "the preferences and resource shares etc." so this project seneds to my computer work whenever should there be a room for it?

O&O (I think I'm evolving into something very special with thses wings)
ID: 10525 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Keck_Komputers
Avatar

Send message
Joined: 17 Sep 05
Posts: 211
Credit: 4,246,150
RAC: 0
Message 10526 - Posted: 7 Feb 2006, 4:31:51 UTC - in response to Message 10525.  

Becasuse, based on your preferences and your resource shares etc. The project should not send you work.


I see

...And based on your knowledge and respect to the science which BOINC resemples, what modification should I do to "the preferences and resource shares etc." so this project seneds to my computer work whenever should there be a room for it?

O&O (I think I'm evolving into something very special with thses wings)

The first setting that usually needs adjusting is the queue length. The maximum recomened queue length is 40% of the shortest deadline divided by the number of projects. The deadlines are commonly 7 days here and that is likely the shortest you have to deal with. It will take approximately 2 times the previous queue length for the client to settle into a groove.

If that doesn't help then increase the resource share for Rosetta. You most likely want to do this in small increments so that you don't have this same problem on a different project.

It also would help if you are set up to run always. Also if possible leave applications in memory.
BOINC WIKI

BOINCing since 2002/12/8
ID: 10526 · Rating: 1 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Paul D. Buck

Send message
Joined: 17 Sep 05
Posts: 815
Credit: 1,812,737
RAC: 0
Message 10527 - Posted: 7 Feb 2006, 6:58:48 UTC

WIthout knowing more about what projects, connect interval, and what the resource shares are on any of them it is not possible for anyone to advise you.

If for example, you had resource shares of 100/100 on two projects, Rosetta@Home and one other, most of the time you would likely have work from both available. BUt, if, for example, RAH did not give you work the other project would continue to run and when RAH did give you work you might not do work for the other project for some interval of time.

As John indicated, changing settings can also cause the system to take a while to get back in the groove. So, if you want us to give a shot at advising you ... we need to know more ...
ID: 10527 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10535 - Posted: 7 Feb 2006, 14:06:31 UTC - in response to Message 10527.  

Becasuse, based on your preferences and your resource shares etc. The project should not send you work.


WIthout knowing more about what projects, connect interval, and what the resource shares are on any of them it is not possible for anyone to advise you....


Thanks Paul.

O&O
ID: 10535 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile O&O
Avatar

Send message
Joined: 11 Dec 05
Posts: 25
Credit: 66,900
RAC: 0
Message 10536 - Posted: 7 Feb 2006, 14:21:26 UTC - in response to Message 10526.  

The first setting that usually needs adjusting is the queue length. The maximum recomened queue length is 40% of the shortest deadline divided by the number of projects. The deadlines are commonly 7 days here and that is likely the shortest you have to deal with. It will take approximately 2 times the previous queue length for the client to settle into a groove.

If that doesn't help then increase the resource share for Rosetta. You most likely want to do this in small increments so that you don't have this same problem on a different project.

It also would help if you are set up to run always. Also if possible leave applications in memory.


Thank you very much KK for your advise.

To begin with, I changed the sitting of the cache on the projects I'm running now (RAH, PG and WCG-FAAH) to 1 day ... each.

I'll wait until the current cache is empty of RAH WUs and act from there with the rest of ... solutions you provided.

Again,... thanks for your time and ... for being here.

O&O
ID: 10536 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : The At-De-At-tach ... concept!



©2024 University of Washington
https://www.bakerlab.org