I notice that the five fresh work units you had when last I looked have now turned into "Client errors" because of the appropriate large data file not being found. For some reason, BOINC couldn't find the appropriate large data files. As a result of this your daily allowance has been cut to 2 which is actually quite fortunate as you can't handle more than that anyway. Your allowance will increase to a maximum of 8 after you start successfully returning results.
Quote:
Which resources shares ??? You mean the work units ???
No. The resource share is the value you can set on the website under the EAH preferences. If you read carefully you will see the words "control resource share and customize graphics" on your account page where you click to get to the EAH preferences. I'm guessing the value will still be "100" which is the default for each project. This will give each project a 50/50 share of your CPU. That share is fine. If you leave your machine on for 10 hours per day, each project would get 5 hours of that.
Quote:
Yes!!! There are the work units files in....
w1_1500.0 8.159kb
l1_1347.0 6.236kb
l1_1278.5 6.239kb
Actually these are mb and not kb. Notice the middle one. Its name tells us it is the large data file that you are currently slicing your work units from. The other two are the large data files associated with the work units showing the client errors I mentioned above. If those were available at the time, I don't understand why you got the client errors saying the data was missing???
Quote:
But at time only this files are in BOINC Manager Window integrated
l1_1347.0__1347.0_0.1_T10_S4lA_3
l1_1347.0__1347.1_0.1_T10_S4lA_1
And these are the "work units" which are essentially small sets of instructions on how to slice up the large data file. Every time you get fresh work, you will get fresh ones of these until the large data file has been "used up".
Quote:
I have change the "work_buf_min_days" value in global_prefs.xml from "0.1" to "2", that was all i had done.
That's not the way you are supposed to do it. It will be overwritten next time BOINC reads your preferences from the website. Please change it there and it will then be properly set.
Quote:
Connect to network about every 0.1 days
This is the value you should change and then BOINC will make the change for you in the .xml file. To stop getting too much EAH work you should set this to about 0.5 - 1.0 days. You can always increase it a bit more later if everything is going well. I think you will find that two days is too much and a lower value will prevent BOINC from going into EDF mode all the time, particularly if you are not running 24/7.
[qoute]
No. The resource share is the value you can set on the website under the EAH preferences. If you read carefully you will see the words "control resource share and customize graphics" on your account page where you click to get to the EAH preferences. I'm guessing the value will still be "100" which is the default for each project. This will give each project a 50/50 share of your CPU. That share is fine. If you leave your machine on for 10 hours per day, each project would get 5 hours of that.
[/qoute]
OK....i checked it now :)
resource share is 100 percent....like the others (SAH and LHC*since yesterday*)
PS: I have post another text but everything goes wrong here...
resource share is 100 percent....like the others (SAH and LHC*since yesterday*)
It is not a percent. All that is cared about is the ratio of the resource shares. 1000 S@H, 100 Einstein will run 10 hours or S@H for every hour of Einstein as will 10 S@H 1 Einstein.
resource share is 100 percent....like the others (SAH and LHC*since yesterday*)
Now that you have three projects with a resource share setting of 100 for each one, each project will be allocated 33.33% of the available cpu time. If you don't leave your computer on 24/7 you may find it difficult to complete the allocated work for all three projects within the deadline, without continually going into EDF mode, if you leave your connect to network setting at 2 days.
When you added LHC to the mix, did you make any adjustments to the preferences on the LHC website? If you did, then the preferences set there will be the most recent and they should propagate to the other projects in time. So, your connect setting may very well now be the value you used when attaching to LHC. With three projects to support, a value of 0.5 days would be very BOINC friendly.
I notice that the five fresh
)
I notice that the five fresh work units you had when last I looked have now turned into "Client errors" because of the appropriate large data file not being found. For some reason, BOINC couldn't find the appropriate large data files. As a result of this your daily allowance has been cut to 2 which is actually quite fortunate as you can't handle more than that anyway. Your allowance will increase to a maximum of 8 after you start successfully returning results.
No. The resource share is the value you can set on the website under the EAH preferences. If you read carefully you will see the words "control resource share and customize graphics" on your account page where you click to get to the EAH preferences. I'm guessing the value will still be "100" which is the default for each project. This will give each project a 50/50 share of your CPU. That share is fine. If you leave your machine on for 10 hours per day, each project would get 5 hours of that.
Actually these are mb and not kb. Notice the middle one. Its name tells us it is the large data file that you are currently slicing your work units from. The other two are the large data files associated with the work units showing the client errors I mentioned above. If those were available at the time, I don't understand why you got the client errors saying the data was missing???
And these are the "work units" which are essentially small sets of instructions on how to slice up the large data file. Every time you get fresh work, you will get fresh ones of these until the large data file has been "used up".
That's not the way you are supposed to do it. It will be overwritten next time BOINC reads your preferences from the website. Please change it there and it will then be properly set.
This is the value you should change and then BOINC will make the change for you in the .xml file. To stop getting too much EAH work you should set this to about 0.5 - 1.0 days. You can always increase it a bit more later if everything is going well. I think you will find that two days is too much and a lower value will prevent BOINC from going into EDF mode all the time, particularly if you are not running 24/7.
Cheers,
Gary.
[qoute] No. The resource
)
[qoute]
No. The resource share is the value you can set on the website under the EAH preferences. If you read carefully you will see the words "control resource share and customize graphics" on your account page where you click to get to the EAH preferences. I'm guessing the value will still be "100" which is the default for each project. This will give each project a 50/50 share of your CPU. That share is fine. If you leave your machine on for 10 hours per day, each project would get 5 hours of that.
[/qoute]
OK....i checked it now :)
resource share is 100 percent....like the others (SAH and LHC*since yesterday*)
PS: I have post another text but everything goes wrong here...
So now a haven't the time to post it again....
regard & cu
BestMischMaker
RE: resource share is 100
)
It is not a percent. All that is cared about is the ratio of the resource shares. 1000 S@H, 100 Einstein will run 10 hours or S@H for every hour of Einstein as will 10 S@H 1 Einstein.
BOINC WIKI
RE: resource share is 100
)
Now that you have three projects with a resource share setting of 100 for each one, each project will be allocated 33.33% of the available cpu time. If you don't leave your computer on 24/7 you may find it difficult to complete the allocated work for all three projects within the deadline, without continually going into EDF mode, if you leave your connect to network setting at 2 days.
When you added LHC to the mix, did you make any adjustments to the preferences on the LHC website? If you did, then the preferences set there will be the most recent and they should propagate to the other projects in time. So, your connect setting may very well now be the value you used when attaching to LHC. With three projects to support, a value of 0.5 days would be very BOINC friendly.
Cheers,
Gary.