I use Mac OSX 10.3.9 with BOINC 5.4.9, and the numbers for the "to completion" under the "tasks" tab are hugely inflated, causing "overcommitted" messages and not downloading work units for some of my projects.
The numbers run down at a rate of maybe a minute (displayed) per 7 seconds (real time), and may start out in the high hundreds of hours for a unit that is done in a day or less.
This seems to cause my computer to have only one or two (of my current 3) projects in que at any given time, so the others won't download until some other workunit finishes. If I don't go online often enough, this leaves my machine without work, waiting to connect.

TIA

Dave Cole

---

I use Mac OSX 10.3.9 with BOINC 5.4.9, and the numbers for the "to completion" under the "tasks" tab are hugely inflated, causing "overcommitted" messages and not downloading work units for some of my projects.
The numbers run down at a rate of maybe a minute (displayed) per 7 seconds (real time), and may start out in the high hundreds of hours for a unit that is done in a day or less.
This seems to cause my computer to have only one or two (of my current 3) projects in que at any given time, so the others won't download until some other workunit finishes. If I don't go online often enough, this leaves my machine without work, waiting to connect.

TIA

Dave Cole

What project?
Can you tell us the duration correction factor for the project? (look on the websites in your account information)

---

Kathryn :o)

What project?
Can you tell us the duration correction factor for the project? (look on the websites in your account information)

I am currently running Predictor, Einstein, and SETI.
SETI seems to get the most accurate numbers and Predictor the worst.
I went to the "your account" page on both Predictor and Einstein and found nothing that looked like "duration correction factor".

Dave

---

What project?
Can you tell us the duration correction factor for the project? (look on the websites in your account information)

I am currently running Predictor, Einstein, and SETI.
SETI seems to get the most accurate numbers and Predictor the worst.
I went to the "your account" page on both Predictor and Einstein and found nothing that looked like "duration correction factor".

Dave

The duration correction factor is found in your account - Computers on this account View - Computer ID number

Einstein has just changed it application from the S4 to S5. S5 take longer to process. A few people have notice that the estimates are somewhat off right now. You may need to crunch a couple of WU before before Boinc can adjust to the new times and correct for the bad estimates.

I haven't paid too much attention to Predictor WU times so I can't say what is going on there.

I find that it is usually best, with multiple projects, to keep your General preference connection time to below 2 days if possible. I keep mine at the default .1

Don't worry too much about Boinc running in EDF mode. With Seti having a wider range of WU crunch times, this will be seen more often. Boinc is designed to handle these situation and is just doing its job as intended.

---

Boinc V 7.4.36
Win7 i5 3.33G 4GB NVidia 470

The duration correction factor is found in your account - Computers on this account View - Computer ID number

Einstein has just changed it application from the S4 to S5. S5 take longer to process. A few people have notice that the estimates are somewhat off right now. You may need to crunch a couple of WU before before Boinc can adjust to the new times and correct for the bad estimates.

I haven't paid too much attention to Predictor WU times so I can't say what is going on there.

I find that it is usually best, with multiple projects, to keep your General
preference connection time to below 2 days if possible. I keep mine at the default .1

Don't worry too much about Boinc running in EDF mode. With Seti having a wider range of WU crunch times, this will be seen more often. Boinc is designed to handle these situation and is just doing its job as intended.

Found them. I still don't know what to do about it.
I reproduced all three below to see if someone can tell me what to do.
I stayed with the default prefs, too.
I have a manual dial-up connection that only connects when I am here.

FROM Predictor:
Operating System Darwin 7.9.0
Memory 384 MB
Cache 976.56 KB
Swap space 0 MB
Total disk space 4.24 GB
Free Disk Space 0.71 GB
Measured floating point speed 201.41 million ops/sec
Measured integer speed 468.03 million ops/sec
Average upload rate 0.13 KB/sec
Average download rate 0.78 KB/sec
Average turnaround time 1.36 days
Maximum daily WU quota per CPU 200/day
Results 15
Number of times client has contacted server 2733
Last time contacted server 3 Jul 2006 1:10:26 UTC
% of time BOINC client is running 86.6416 %
While BOINC running, % of time work is allowed 99.8289 %
Average CPU efficiency 0.634136
Result duration correction factor 18.999474

FROM Einstein:
Operating System Darwin 7.9.0
Memory 384 MB
Cache 976.56 KB
Swap space 0 MB
Total disk space 4.24 GB
Free Disk Space 0.72 GB
Measured floating point speed 201.41 million ops/sec
Measured integer speed 468.03 million ops/sec
Average upload rate 0.14 KB/sec
Average download rate 0.91 KB/sec
Average turnaround time 9.38 days
Maximum daily WU quota per CPU 32/day
Results 1
Number of times client has contacted server 1371
Last time contacted server 30 Jun 2006 19:22:05 UTC
% of time BOINC client is running 85.6052 %
While BOINC running, % of time work is allowed 99.8156 %
Average CPU efficiency 0.663343
Result duration correction factor 19.757044

FROM SETI:
Operating System Darwin 7.9.0
Memory 384 MB
Cache 976.56 KB
Swap space 0 MB
Total disk space 4.24 GB
Free Disk Space 0.72 GB
Measured floating point speed 201.41 million ops/sec
Measured integer speed 468.03 million ops/sec
Average upload rate 0.14 KB/sec
Average download rate 8.18 KB/sec
Average turnaround time 8.81 days
Maximum daily WU quota per CPU 100/day
Results 1
Number of times client has contacted server 644
Last time contacted server 1 Jul 2006 3:16:15 UTC
% of time BOINC client is running 85.7623 %
While BOINC running, % of time work is allowed 99.8176 %
Average CPU efficiency 0.651837
Result duration correction factor 1

TIA

Dave

---

The duration factor for Predictor 18.999474 and Einstein 19.757044 are way out of whack, while Seti is still a perfect 1. The duration factor is meant to calibrate the completion estimate supplied by the project to dispay the amount of time your computer takes. In your case it is producing an over blown number.

Why it happened, I can't explain. What you can do about it, not much other than wait. You could edit your client_state.xml file and change the numbers, but that can be tricky if you don't know exactly what your doing.

I suggest just leaving it. Boinc will slowly bring the value back down to where it should be with each WU you crunch. You should notice that the next WU you get will be a bit closer to the correct estimate.

Boinc was design to increase the factor rapidly to avoid downloading too much work. It also is meant to reduce the number slowly for the same reason.

Patience is the key in this case.

---

Boinc V 7.4.36
Win7 i5 3.33G 4GB NVidia 470

The duration factor for Predictor 18.999474 and Einstein 19.757044 are way out of whack, while Seti is still a perfect 1. The duration factor is meant to calibrate the completion estimate supplied by the project to dispay the amount of time your computer takes. In your case it is producing an over blown number.

Why it happened, I can't explain. What you can do about it, not much other than wait. You could edit your client_state.xml file and change the numbers, but that can be tricky if you don't know exactly what your doing.

I suggest just leaving it. Boinc will slowly bring the value back down to where it should be with each WU you crunch. You should notice that the next WU you get will be a bit closer to the correct estimate.

Boinc was design to increase the factor rapidly to avoid downloading too much work. It also is meant to reduce the number slowly for the same reason.

Patience is the key in this case.

Thank you very much A.B.
Where is this "client_state.xml" file and *how* tricky is editing it?
I am fairly good at following instructions, but I can't do much in new territory without a map and compass.
Is this stored on my machine or on the servers?
-------
Oh, wait, I just found it as a text file on my drive under BOINC.
I'll try messin' with it and see what happens.
Wish me luck.

Thanks again!

Dave

---

Thank you very much A.B.
Where is this "client_state.xml" file and *how* tricky is editing it?
I am fairly good at following instructions, but I can't do much in new territory without a map and compass.
Is this stored on my machine or on the servers?
-------
Oh, wait, I just found it as a text file on my drive under BOINC.
I'll try messin' with it and see what happens.
Wish me luck.

Thanks again!

Dave

OK, you seem the adventurous type. So here's your road map. It's not overly complicated, but could cause Boinc to become confused if not done right.

A note before starting. This file contains details about your accounts and acts pretty well as Boinc's road map to the projects your attached to and the state of WU you are crunching. Great care is required to not mess it up. Don't use the 'Save as' from the text editor as you may loose some of the formatting characters that the text editor doesn't recognize.

First, set all the projects to 'no new work/tasks' in the Projects tab and exit Boinc completely so that it isn't crunching anymore. I also suggest making an extra copy of the 'client_state.xml' file on your desktop using the drag an drop method. That way if things get messed up you can restore it.

Open the 'client_state.xml' in the Boinc folder with your text editor.

Find the section for Einstein with the header
<project>
<master_url>http://einstein.phys.uwm.edu/</master_url>
<project_name>Einstein@Home</project_name>
.
about 20 line lower you will see
.
<duration_correction_factor>19.757044</duration_correction_factor>
change the number to 1

Now locate Predictor and do the same thing
<project>
<master_url>http://predictor.scripps.edu/</master_url>
<project_name>Predictor @ Home</project_name>
.
.
<duration_correction_factor>18.999474</duration_correction_factor>
change the number to 1

The actual numbers may have changed since your post.
Save the file (not Save as).

Restart Boinc and allow it to crunch for a while before allowing each project (Project tab) to fetch new work again.

A final note:
If you changed your General preference 'Connect to network about every' to a large number you may want to reduce it, as Boinc may try to get more work than it can handle.

---

Boinc V 7.4.36
Win7 i5 3.33G 4GB NVidia 470

OK, you seem the adventurous type. So here's your road map. It's not overly complicated, but could cause Boinc to become confused if not done right.

A note before starting. This file contains details about your accounts and acts pretty well as Boinc's road map to the projects your attached to and the state of WU you are crunching. Great care is required to not mess it up. Don't use the 'Save as' from the text editor as you may loose some of the formatting characters that the text editor doesn't recognize.

First, set all the projects to 'no new work/tasks' in the Projects tab and exit Boinc completely so that it isn't crunching anymore. I also suggest making an extra copy of the 'client_state.xml' file on your desktop using the drag an drop method. That way if things get messed up you can restore it.

Open the 'client_state.xml' in the Boinc folder with your text editor.

Find the section for Einstein with the header
<project>
<master_url>http://einstein.phys.uwm.edu/</master_url>
<project_name>Einstein@Home</project_name>
.
about 20 line lower you will see
.
<duration_correction_factor>19.757044</duration_correction_factor>
change the number to 1

Now locate Predictor and do the same thing
<project>
<master_url>http://predictor.scripps.edu/</master_url>
<project_name>Predictor @ Home</project_name>
.
.
<duration_correction_factor>18.999474</duration_correction_factor>
change the number to 1

The actual numbers may have changed since your post.
Save the file (not Save as).

Restart Boinc and allow it to crunch for a while before allowing each project (Project tab) to fetch new work again.

A final note:
If you changed your General preference 'Connect to network about every' to a large number you may want to reduce it, as Boinc may try to get more work than it can handle.

Thank you, A.B.
It seems to have worked well.

Dave

---