The last change made was to chance the % load from 100 for all projects to 130 for Einstein. I then left over nite before I finially got some work units.
I am getting different error messages from both LHC and Einstein
Einstein@Home|Deferring communication with project for 7 hours
Quote:
ProteinPredictorAtHome is now sendig data.
Einstein is still NOT sending data out.
Quote:
Thanks for the network tip.
*Seti >> is not sending work.
*ProteinPredictorAtHome >> is not sending out work.
*climateprediction >> is still working (but it is hard to tell if new units are being sent out) but only on my single CPU machine
*Einstein >> is the only one not sending work that claims to be working
*LHC >> working and now the only work units on my Dual CPU machine.
Quote:
Hello Louis!
It's caused by the long term debt, and if Einstein is considered by Boinc to have done enough lately, and the other projects have to have it's share for the next time, that's wor as designed. Einstein is likely to be a candidate for this behaviour, as it's short deadlines and long crunchtimes force Boinc more often in "panic mode" (crunch imediately, because deadline looms). Afterwards it's given some time to relax.
The time share is no longer a strict daily quota since introduction of this feature.
You have two possibilities to overcome this from resulting in getting no WUs:
1. Get your value for "Connect to network every...days" in the preferences up to some days, resulting in a bigger WU - cache.
2. Only in emergency! Reset the project. Everything not yet crunched or reported will get lost, so don't do it while some WUs are still sitting in there. But the long term debt will get lost as well.
BTW:
Try to avoid writing only in capitals, as it's considered as shouting in fora.
I am getting different error messages from both LHC and Einstein
Einstein@Home|Deferring communication with project for 7 hours
Try following:
go to the tab "Projects", hit the "Update" button and have a look at the tab "Messages" to see what happens. Don't hit the same "Update" button again if you see in the message something like "last RPC too recent". In that case leave Boinc for a while like it is.
FYI: no kidding, I just received 4 wu's (along with 4 s@h and 3 lhc) as soon as I connected this evening (had been running disconnected since reset last night) ... guess it was time for me to 'reset' the e@h project - never had to do that before?! So, in other words, resetting the project looks like it's fixed my issue of no work. However, my computer is obviously 'over committed' at the moment and in panic mode.
I also modified my e@h and/or general boinc settings today to connect every 2.5 days (previously was 2), shares at: 35% s@h, 35% e@h, and 30% lhc, and detached from e@h on all other computers [ for now :( ]
A standard formula for maximum "connect to" time is take the minimum return time (in your case 7 days from E@H) and divide by the number of projects (in your case 3). So for y ou, it's 2.3 If you go above this number you will likely go into Earliest deadline on Einstien, which will prevent you from down loading units from time to time while the long term debt is worked off.
Unless you have MAJOR problems with BOINC there is almost NEVER a reason to reset everything.
guess it was time for me to 'reset' the e@h project - never had to do that before?! So, in other words, resetting the project looks like it's fixed my issue of no work.
Resetting is only exacerbating your problems. Just let your Long Term Debt work itself out. If you keep resetting, you aren't allowing BOINC to honor your resource share.
I run 50/50 E.-&-S.@home
Einstien always wants to send me 0 sec work
After it runs the last stashed
So what I do and damned if it doesn't work is:
Shut down 4.45 an then simply click the exe for 4.19
4.19 comes up and connects and updates and runs the benchmarks
And I get about the same amount of WU that are there for Seti
From Einstien-
When it's up and cooking I shut it down and hit the exe for 4.45-
4.45 installs, runs benchmarks, and goes to work
With all the new WU.s showing up in the work tab-
I do this every 2 weeks or so
I'm not prepared to answer any questions as to why this works...cc
I'm not prepared to answer any questions as to why this works...cc
I am... 4.19 uses the old scheduling algorithm. By doing what you describe, you are in effect making BOINC not honor your resource share.
If you really feel you must sneak around the back of the scheduler, you can simply reset the project multiple times, (make sure you have no results completed or ready to send) or download BOINC Debt Viewer and reset your Long Term Debt.
Personally, I would just leave it alone and let it do what it's supposed to do...
Now I have 16 WU's from
)
Now I have 16 WU's from Einstein and 12 from LHC
The last change made was to chance the % load from 100 for all projects to 130 for Einstein. I then left over nite before I finially got some work units.
I am getting different error messages from both LHC and Einstein
Einstein@Home|Deferring communication with project for 7 hours
RE: I am getting different
)
Try following:
go to the tab "Projects", hit the "Update" button and have a look at the tab "Messages" to see what happens. Don't hit the same "Update" button again if you see in the message something like "last RPC too recent". In that case leave Boinc for a while like it is.
Greetings from Belgium
Thierry
FYI: no kidding, I just
)
FYI: no kidding, I just received 4 wu's (along with 4 s@h and 3 lhc) as soon as I connected this evening (had been running disconnected since reset last night) ... guess it was time for me to 'reset' the e@h project - never had to do that before?! So, in other words, resetting the project looks like it's fixed my issue of no work. However, my computer is obviously 'over committed' at the moment and in panic mode.
I also modified my e@h and/or general boinc settings today to connect every 2.5 days (previously was 2), shares at: 35% s@h, 35% e@h, and 30% lhc, and detached from e@h on all other computers [ for now :( ]
A standard formula for
)
A standard formula for maximum "connect to" time is take the minimum return time (in your case 7 days from E@H) and divide by the number of projects (in your case 3). So for y ou, it's 2.3 If you go above this number you will likely go into Earliest deadline on Einstien, which will prevent you from down loading units from time to time while the long term debt is worked off.
Unless you have MAJOR problems with BOINC there is almost NEVER a reason to reset everything.
RE: guess it was time for
)
Resetting is only exacerbating your problems. Just let your Long Term Debt work itself out. If you keep resetting, you aren't allowing BOINC to honor your resource share.
Awesome explanations ...
)
Awesome explanations ... thanks! Learn something new every day!
I run 50/50
)
I run 50/50 E.-&-S.@home
Einstien always wants to send me 0 sec work
After it runs the last stashed
So what I do and damned if it doesn't work is:
Shut down 4.45 an then simply click the exe for 4.19
4.19 comes up and connects and updates and runs the benchmarks
And I get about the same amount of WU that are there for Seti
From Einstien-
When it's up and cooking I shut it down and hit the exe for 4.45-
4.45 installs, runs benchmarks, and goes to work
With all the new WU.s showing up in the work tab-
I do this every 2 weeks or so
I'm not prepared to answer any questions as to why this works...cc
RE: I'm not prepared to
)
I am... 4.19 uses the old scheduling algorithm. By doing what you describe, you are in effect making BOINC not honor your resource share.
If you really feel you must sneak around the back of the scheduler, you can simply reset the project multiple times, (make sure you have no results completed or ready to send) or download BOINC Debt Viewer and reset your Long Term Debt.
Personally, I would just leave it alone and let it do what it's supposed to do...
>>>Personally, I would just
)
>>>Personally, I would just leave it alone and let it do what it's supposed to do...
Thanks for this explaination.
I will try to reset long term debt and see if that cures it...cc