Requesting '(seconds)' seconds of new work, and reporting '(count)' results

From Unofficial BOINC Wiki

Jump to: navigation, search

Contents

[edit] General

Message Type: Information Message

This message means that the BOINC Client Software needs to obtain additional Work Units to fill its Work Buffer so you will not "run dry" and your computer is always going to be busy. The number of seconds that it is going to request will (or should) match up to the number of days of work that you desire and has been set in your Preference Settings.

The BOINC Client Software will attempt to obtain work from the Project where the Long Term Debt is the highest. Failing that, it will go to the Project that has the next highest Long Term Debt. It will also attempt to keep an interesting assortment of work on your local client.

This message is changed from the earlier form in that it now also informs you of the number of Results that are also being returned within this Scheduler Request Message.

Note:
A Scheduler Request Message for one second's worth of work is equivalent to asking for one Work Unit.


[edit] Version Information

[edit] Alternate Message Formats

  • Requesting "x" seconds of work
Obsolete/Changed in Version: 4.45
  • Requesting '(seconds)' seconds of new work, and reporting '(count)' results
New in Version 5..x

[edit] Example Log(s)

[edit] Scheduler Request Succeeded - Reason: Requested By User

(1) 7/8/2005 11:46:55 AM|resourcemeasurement|Sending scheduler request to
                                             http://winerror.cs.berkeley.edu/resourcemeasurement_cgi/cgi
(2) 7/8/2005 11:46:55 AM|resourcemeasurement|Reason: Requested by user
(3) 7/8/2005 11:46:55 AM|resourcemeasurement|Requesting 4320 seconds of work, returning 0 results
(4) 7/8/2005 11:46:57 AM|resourcemeasurement|Scheduler request to
                                             http://winerror.cs.berkeley.edu/resourcemeasurement_cgi/cgi
                                             succeeded
(5) 7/8/2005 11:47:58 AM|resourcemeasurement|Fetching master file
(6) 7/8/2005 11:47:59 AM|resourcemeasurement|Master page download succeeded

Line-By-Line Explanation

  1. Sending scheduler request to '(url)'
    Initiating the communication session with the Project's Scheduler.
  2. Reason: Requested by user
    Because the Participant wanted to contact this Project.
  3. Requesting '(seconds)' seconds of work, returning '(count)' results
    Within the Scheduler Request Message we are asking for work to the tune of 'x' seconds, and also returning 'count' Results to the Project. In this case, there was a need a small amount of additional work, so the time was set to 4320 seconds worth of work needed; and there are no Results to "report".
  4. Scheduler request to '(url)' succeeded
    And we had success contacting the Scheduler!
  5. Fetching master file
    We had going to use the Project's Master URL to fetch the Master File which will have address (URL) information embedded within that will allow us to contact the Project's Schedulers.
  6. Master page download succeeded
    And we were successful in obtaining the Master Page.

[edit] Scheduler Request - 0 Seconds & 0 Results

This example means that you are running 4.72 or later, and one or more of the following conditions is true:

  1. There is too much work on your machine already, and downloading any more will possibly run something past the Result Deadline.
  2. The machine has enough work in the Work Buffer, and specifically the project has enough work on this machine.
  3. The project has had to borrow some CPU Time to make a Result Deadline. It is paying the other projects back by not downloading anything for a while on this machine.
(1) 9/20/2005 8:26:33 PM|XtremLab|Sending scheduler request to 
                                  http://xw01.lri.fr:4320/xtremlab_cgi/cgi
(2) 9/20/2005 8:26:33 PM|XtremLab|Requesting 0 seconds of work, returning 0 results
(3) 9/20/2005 8:26:34 PM|XtremLab|Scheduler request to 
                                  http://xw01.lri.fr:4320/xtremlab_cgi/cgi succeeded

Line-By-Line Explanation

  1. Sending scheduler request to '(url)'
    • We are going to send in a report to the Project.
  2. Requesting '(seconds)' seconds of work, returning '(count)' results
    • And we are not asking for more work, nor are we "reporting" a Result.
  3. Scheduler request to '(url)' succeeded
    • And we were successful.

[edit] Scheduler Request - Trickle Reporting

(1) 2005-09-07 02:53:51 [climateprediction.net] Sending scheduler request to
                                                http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi
(2) 2005-09-07 02:53:51 [climateprediction.net] Requesting 0 seconds of work, returning 0 results
(3) 2005-09-07 02:53:52 [climateprediction.net] Scheduler request to
                                                http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi
                                                succeeded

Line-By-Line Explanation

  1. Sending scheduler request to '(url)'
  2. Requesting '(seconds)' seconds of work, returning '(count)' results
    • And we are not asking for more work, nor are we "reporting" a Result.
  3. Scheduler request to '(url)' succeeded
    • And we were successful.

[edit] Outdated Code Verification Key

6/19/2005 12:05:00 AM|SETI@home|Sending scheduler request to
                                http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
6/19/2005 12:05:00 AM|SETI@home|Requesting 1 seconds of work, returning 0 results
6/19/2005 12:05:01 AM|SETI@home|Scheduler request to
                                http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
6/19/2005 12:05:01 AM|SETI@home|Message from server: You may have an outdated code verification key.
                                This may prevent you from accepting new executables.
                                If the problem persists, detach/attach the project.
6/19/2005 12:05:02 AM|SETI@home|Started download of 29fe04aa.15717.8802.148560.195
6/19/2005 12:05:06 AM|SETI@home|Finished download of 29fe04aa.15717.8802.148560.195
6/19/2005 12:05:06 AM|SETI@home|Throughput 98162 bytes/sec
6/19/2005 12:05:06 AM|         |request_reschedule_cpus: files downloaded

Line-By-Line Explanation

Initiating the communication session with the Project's Scheduler.
Within the Scheduler Request Message we are asking for work to the tune of 'x' seconds, and also returning 'count' Results to the Project. In this case, there was a need a small amount of additional work, so the time was set to 1 seconds worth of work needed; and there are no Results to "report".
And we had success contacthing the Scheduler!
During the communication session with the Scheduler, the Scheduler is warning that the Code Verification Key may be out of date or corrupted. However, ... this does not stop us from getting new work issued to us in the form of one additional Work Unit.
This message indicates the "File Name" of the file that is to be downloaded, and that the download operation has been started.
This message indicates that the file has completed the download operation.
This message indicates the measured speed of the file transfer.
This message indicates that the Work Scheduler needs to re-examine the current schedule to see if it needs to be changed based on the fact that the contents of the Work Buffer has been changed.

[edit] Scheduler Request Failed - Timeout (-182)

7/8/2005 10:03:31 AM|ProteinPredictorAtHome|Sending scheduler request to
                                            http://predictor.scripps.edu/predictor_cgi/cgi
7/8/2005 10:03:31 AM|ProteinPredictorAtHome|Reason: To fetch work
7/8/2005 10:03:31 AM|ProteinPredictorAtHome|Requesting 4320 seconds of work, returning 0 results
7/8/2005 10:13:31 AM|ProteinPredictorAtHome|Scheduler request to
                                            http://predictor.scripps.edu/predictor_cgi/cgi
                                            failed with a return value of -182
7/8/2005 10:13:31 AM|ProteinPredictorAtHome|No schedulers responded
7/8/2005 10:13:32 AM|ProteinPredictorAtHome|Deferring communication with project for 1 minutes
                                            and 55 seconds

Line-By-Line Explanation

Starting a communication session with the Project's Scheduler.
Te get more work ...
Asking for more work without returning any Results.
But we failed to complete the request because too much time passed before we got an answer.
Because no Scheduler responded to the Scheduler Request Message.
So, we wait a period of time based on Exponential Back-Off, each attempt is tried after a longer and longer delay.

[edit] Requesting Work - Returning Results

(1) 6/8/2005 1:58:41 PM|SETI@home|Sending scheduler request to
                                  http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
(2) 6/8/2005 1:58:41 PM|SETI@home|Requesting 0 seconds of work, returning 3 results

Line-By-Line Explanation

  1. Sending scheduler request to '(url)'
  2. Requesting '(seconds)' seconds of work, returning '(count)' results
    • Within the Scheduler Request Message we are asking for work to the tune of 'x' seconds, and also returning 'count' Results to the Project. In this case, there was no need for additional work, so the time was set to 0 seconds worth of work needed.

[edit] Other Related Messages

  • None.


[edit] Developer Information

Code Location:   boinc/client/scheduler_op.C
Function:   SCHEDULER_OP::start_rpc(PROJECT* p)
Documentation:   boinc/client/scheduler_op.C
CVS Extract:   cvs -d :pserver:anonymous:@alien.ssl.berkeley.edu:/home/cvs/cvsroot checkout boinc/client/scheduler_op.C
Personal tools
RSS Feeds
BOINC Wiki RSS feeds RSS Feeds
Powered by BOINC!
Powered by BOINC