Deprecated: Function get_magic_quotes_gpc() is deprecated in /disks/centurion/b/carolyn/b/home/boincadm/projects/beta/html/inc/util.inc on line 663
Run times ??

Run times ??

Message boards : SETI@home Enhanced : Run times ??
Message board moderation

To post messages, you must log in.

AuthorMessage
Dave Stegner
Volunteer tester
Avatar

Send message
Joined: 23 Aug 11
Posts: 16
Credit: 74,238
RAC: 0
United States
Message 41195 - Posted: 25 Aug 2011, 23:32:31 UTC
Last modified: 25 Aug 2011, 23:51:08 UTC

I am new to Seti Beta. I loaded 3 of my machines last week.

I have completed and validated a number of tasks using standard v7 app.

Before a wu starts the estimated run tile is 1 day and prox 1 hour, all 3 machines. After a wu starts, it immediately drops to 8 hours and completes in 8 hours.

Why are the remaining wu's not recalculating to 8 hours??

Dave
ID: 41195 · Report as offensive
Josef W. Segur
Volunteer tester

Send message
Joined: 14 Oct 05
Posts: 1137
Credit: 1,848,733
RAC: 0
United States
Message 41204 - Posted: 27 Aug 2011, 1:46:36 UTC - in response to Message 41195.  

I am new to Seti Beta. I loaded 3 of my machines last week.

I have completed and validated a number of tasks using standard v7 app.

Before a wu starts the estimated run tile is 1 day and prox 1 hour, all 3 machines. After a wu starts, it immediately drops to 8 hours and completes in 8 hours.

Why are the remaining wu's not recalculating to 8 hours??

Dave

Because the BOINC core client adjusts its duration correction factor (DCF) downward slowly.

Two of your hosts have passed the magic 10 completed/validated mark and the third is close. The server-side adjustments will be in effect for new work after that point, which ought to help. There will probably be an offset in DCF which will work itself out within a day or two.

OTOH, the file from which the v7 work was split has many VLAR and VHAR tasks, the midrange tasks like you got originally for your 3 hosts are going to be rare until we get to the same zone near the end of the current channel. The changing characteristics of the work will probably cause DCF to bounce around somewhat.

There could be more effects going on, I have no experience with BOINC 6.12.33.
                                                                 Joe
ID: 41204 · Report as offensive
uBronan
Volunteer tester
Avatar

Send message
Joined: 29 Jun 07
Posts: 32
Credit: 72,650
RAC: 0
Antarctica
Message 41206 - Posted: 27 Aug 2011, 8:02:32 UTC

The version of boinc does not matter so much but your on the right track the estimate does simply need longer time to get set close to the time needed.
But on some projects is fails miserable, depending higly on the units you receive.

I have projects running where a estimate is really a joke since every unit runs totally different times.

Some do finish after 45 minutes but i have units which run over 90 hours from the same batch, so my point is on this do not weigh heavy on estimates.

Now do not get me wrong some projects you know exactly how long the units will/should run.

For instance on almost all gpu projects i almost know exactly how long they run in minutes just by seeing the name of the unit being processed.

In time you probably make a better estimate then boinc does, if you run it long enough ;).
ID: 41206 · Report as offensive

Message boards : SETI@home Enhanced : Run times ??


 
©2023 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.