No problem (g). I had a problem, didn't know if it was due to boinc scheduling or what causing an inability to keep my machine gainfully employed. There are alot of people much more knowledgeable than I so I just post a question/problem to the hive mind and see what happens.....p
No problem (g). I had a problem, didn't know if it was due to boinc scheduling or what causing an inability to keep my machine gainfully employed. There are alot of people much more knowledgeable than I so I just post a question/problem to the hive mind and see what happens.....p
The "real" problem is that you do not have a sufficient "spread" of projects to handle the situation when one project is off-line.
As you note, there can be problems when one project is down keeping the computers busy. My recommendation is always to have at least 3 projects, more prefferably 5. This has the side effect that you have to be attached to all of the current production projects. However, even with only 3, if you chose CPDN and it ran acceptably well, you would have none of these worries.
The only defect at the moment with CPDN is that with the new Sulfur Model the run time has gone up considerably and the Result Deadline has come down at the same time due to project needs. However, they are working on issuing the older models to those machines that have lower capabiity though the best way to limit to the older models seems to be to change the disk allocation ... also, they have just gone off-line for a week ... :(
For steady stand-by work issue the only other project that may suit would be Predictor@Home.
Hello Paul. I thought I had a 'sufficient' spread of projects. If one is down, the other takes up the slack. But with einstein not loading four days of work if it had one w/u left then that caused my machine to sit idle, which is not what I see as one of the primary concerns of boinc. Without boinc, I could load and run a seti cruncher and if something went wrong, then I could just load and run an einstein until seti came back. Boinc is just the wrapper to make things automatic...and my point of contention is that boinc should have allowed einstein to load more w/u's..just a tempest in a teapot...p
Hello Paul. I thought I had a 'sufficient' spread of projects. ... ... Boinc is just the wrapper to make things automatic...and my point of contention is that boinc should have allowed einstein to load more w/u's..just a tempest in a teapot...p
Um, of course.
I was pretty sure I had that understanding. I was just commenting. It goes along with the last interview of Dr. Anderson ... BOINC is not designed or intended for "x" ... as it were.
At any rate, SETI@Home seems to be back up again and running smoothly, more or less, so, the immediate problem of yours should be solved now. There have been some minor issues with the CPU Scheduler and Work Scheduler on the Client-side software. But, we won't see all of those till the 5.x.y version comes out. I find 4.45 works well (with a couple minor issues, like reporting results immediately after the upload instead of waiting) and so does 4.72 OS-X version. BUt, am eager for the 5.x.y to be issued.
Won't you be able to pay off the long term debt as slowly or fast as you want to then? Or will the long term debt buildup actually get worse if you keep on switching other projects on and off?
You can get wherever you want micro-managing your BOINC. But I don't think this is feasible for most of users. I think BOINC should behave well enough to just let it do its own job and just check from time to time wether everything still runs OK.
For me, BOINC just plain works as intended. Thing is that I'd like it behave slightly differently, that's all.
I can't get any work from SETI nor Einstein, so as Lattice and Orbit are not active yet, I am really running only Protein Predictor. I detatched form CPDN and got LHC on no new work and will probably detach due to the really short deadlines. I have no queued WUs at all. Could you plese explain me this? Thanks!
Einstein
short term debt 0.000000
long term debt -24916.400739
share 100.000000
SETI
short term debt 0.000000
long term debt -82035.314387
share 100.000000
I can't get any work from SETI nor Einstein, so as Lattice and Orbit are not active yet, I am really running only Protein Predictor. I detatched form CPDN and got LHC on no new work and will probably detach due to the really short deadlines. I have no queued WUs at all. Could you plese explain me this? Thanks!
Einstein
short term debt 0.000000
long term debt -24916.400739
share 100.000000
SETI
short term debt 0.000000
long term debt -82035.314387
share 100.000000
"No New Work" doesn't remove a project from the work fetch calculation, so your negative Long Term Debt for SETI and Einstein are saying "no thanks" because of LHC and Predictor. Until you've done enough Predictor work to "catch up" with SETI and Einstein, you won't see any WUs from them. If you "Suspend" LHC instead of "No New Work" it, it will help a bit, but you're still going to run Predictor only for a while unless you change resource shares or in some other way micro-manage things. This is normal. Predictor sometimes HAS no work to send out. If you still have no SETI or Einstein work and have no Predictor WUs, "Suspend" Predictor for a few minutes and it will hopefully straighten out.
I upset a few people over in seti with a discussion about this topic. As Metod notes that "most" people do not understand every item in the client_state.xml and editing the wrong item could be very bad for your already high frustration level...
There is a 3rd Party Tool BoincDv that is designed to only look at LTD (there are other 3rd party Tools there also). It has a simple button called "Clear Debts." If your Resource Share is not correct then LTD will whack out again...
So the best advice I can give is run things down to Zero. Suspend your projects from BOINCMgr then Stop BOINC and Then make a Backup Copy of your BOINC folder. After that use the BoincDV tool to reset the LTD. As You restart BOINC Everyone will start clamering for Work... This will probably restart the spiral down... So it you have two project and one is down only get work from the running one. Tell BOINCMgr No New Work, and allow it time to reduce the work queue and then Enable the Suspended Project. It will have had a bit of time to start catching up with what is happening
Quote:
I can't get any work from SETI nor Einstein, so as Lattice and Orbit are not active yet, I am really running only Protein Predictor. I detatched form CPDN and got LHC on no new work and will probably detach due to the really short deadlines. I have no queued WUs at all. Could you plese explain me this? Thanks!
Einstein
short term debt 0.000000
long term debt -24916.400739
share 100.000000
SETI
short term debt 0.000000
long term debt -82035.314387
share 100.000000
There is a 3rd Party Tool BoincDv that is designed to only look at LTD (there are other 3rd party Tools there also). It has a simple button called "Clear Debts." If your Resource Share is not correct then LTD will whack out again...
Unfortunately for me this long-term debt adjusting tool is only designed for Boinc 4.4x. I am running 4.72.... why does this always happen like optimized seti application but none for the other Boinc projects.... I know open source code... sigh
No problem (g). I had a
)
No problem (g). I had a problem, didn't know if it was due to boinc scheduling or what causing an inability to keep my machine gainfully employed. There are alot of people much more knowledgeable than I so I just post a question/problem to the hive mind and see what happens.....p
RE: No problem (g). I had
)
The "real" problem is that you do not have a sufficient "spread" of projects to handle the situation when one project is off-line.
As you note, there can be problems when one project is down keeping the computers busy. My recommendation is always to have at least 3 projects, more prefferably 5. This has the side effect that you have to be attached to all of the current production projects. However, even with only 3, if you chose CPDN and it ran acceptably well, you would have none of these worries.
The only defect at the moment with CPDN is that with the new Sulfur Model the run time has gone up considerably and the Result Deadline has come down at the same time due to project needs. However, they are working on issuing the older models to those machines that have lower capabiity though the best way to limit to the older models seems to be to change the disk allocation ... also, they have just gone off-line for a week ... :(
For steady stand-by work issue the only other project that may suit would be Predictor@Home.
Hello Paul. I thought I had
)
Hello Paul. I thought I had a 'sufficient' spread of projects. If one is down, the other takes up the slack. But with einstein not loading four days of work if it had one w/u left then that caused my machine to sit idle, which is not what I see as one of the primary concerns of boinc. Without boinc, I could load and run a seti cruncher and if something went wrong, then I could just load and run an einstein until seti came back. Boinc is just the wrapper to make things automatic...and my point of contention is that boinc should have allowed einstein to load more w/u's..just a tempest in a teapot...p
RE: Hello Paul. I thought
)
Um, of course.
I was pretty sure I had that understanding. I was just commenting. It goes along with the last interview of Dr. Anderson ... BOINC is not designed or intended for "x" ... as it were.
At any rate, SETI@Home seems to be back up again and running smoothly, more or less, so, the immediate problem of yours should be solved now. There have been some minor issues with the CPU Scheduler and Work Scheduler on the Client-side software. But, we won't see all of those till the 5.x.y version comes out. I find 4.45 works well (with a couple minor issues, like reporting results immediately after the upload instead of waiting) and so does 4.72 OS-X version. BUt, am eager for the 5.x.y to be issued.
RE: Won't you be able to
)
You can get wherever you want micro-managing your BOINC. But I don't think this is feasible for most of users. I think BOINC should behave well enough to just let it do its own job and just check from time to time wether everything still runs OK.
For me, BOINC just plain works as intended. Thing is that I'd like it behave slightly differently, that's all.
Metod ...
I can't get any work from
)
I can't get any work from SETI nor Einstein, so as Lattice and Orbit are not active yet, I am really running only Protein Predictor. I detatched form CPDN and got LHC on no new work and will probably detach due to the really short deadlines. I have no queued WUs at all. Could you plese explain me this? Thanks!
Einstein
short term debt 0.000000
long term debt -24916.400739
share 100.000000
SETI
short term debt 0.000000
long term debt -82035.314387
share 100.000000
***********************************
WinXP Home, P4 3.2GHz HT, 512Mb RAM. Running SETI, Predictor, LHC, Einstein, Orbit, Lattice and Folding
RE: I can't get any work
)
"No New Work" doesn't remove a project from the work fetch calculation, so your negative Long Term Debt for SETI and Einstein are saying "no thanks" because of LHC and Predictor. Until you've done enough Predictor work to "catch up" with SETI and Einstein, you won't see any WUs from them. If you "Suspend" LHC instead of "No New Work" it, it will help a bit, but you're still going to run Predictor only for a while unless you change resource shares or in some other way micro-manage things. This is normal. Predictor sometimes HAS no work to send out. If you still have no SETI or Einstein work and have no Predictor WUs, "Suspend" Predictor for a few minutes and it will hopefully straighten out.
atotos I upset a few
)
atotos
I upset a few people over in seti with a discussion about this topic. As Metod notes that "most" people do not understand every item in the client_state.xml and editing the wrong item could be very bad for your already high frustration level...
There is a 3rd Party Tool BoincDv that is designed to only look at LTD (there are other 3rd party Tools there also). It has a simple button called "Clear Debts." If your Resource Share is not correct then LTD will whack out again...
So the best advice I can give is run things down to Zero. Suspend your projects from BOINCMgr then Stop BOINC and Then make a Backup Copy of your BOINC folder. After that use the BoincDV tool to reset the LTD. As You restart BOINC Everyone will start clamering for Work... This will probably restart the spiral down... So it you have two project and one is down only get work from the running one. Tell BOINCMgr No New Work, and allow it time to reduce the work queue and then Enable the Suspended Project. It will have had a bit of time to start catching up with what is happening
R/
Al
There is a 3rd Party Tool
)
There is a 3rd Party Tool BoincDv that is designed to only look at LTD (there are other 3rd party Tools there also). It has a simple button called "Clear Debts." If your Resource Share is not correct then LTD will whack out again...
Unfortunately for me this long-term debt adjusting tool is only designed for Boinc 4.4x. I am running 4.72.... why does this always happen like optimized seti application but none for the other Boinc projects.... I know open source code... sigh
Thanks! With BoincDV I was
)
Thanks! With BoincDV I was able to get a new Einstein WU........ Very happy me! ;-D
***********************************
WinXP Home, P4 3.2GHz HT, 512Mb RAM. Running SETI, Predictor, LHC, Einstein, Orbit, Lattice and Folding