No problem with the credit claim for a S@H enhanced WU about the same as with C3r 5.2.13, but I´ve heard that if you use Trux you skuold diaable it for S@H enhanced since tx36 dosen´t support that onte just keeping it for S@H standard if it´s possible.
I`m getteing a little bit higher claimed credit with 5.5.0 then I got with C3r client 5.2.13. I think because for some reason I got little better benchmark. I now claim about 22 credit instead of 16 for a WU, I´m also using akosf S41.07 optimezed einstein at home client. About Seti at Home I have to wait to see, no reported WU their yet.
That is an amazing jump up ! *smile* I hope that happens to me as well.
seems as if this is the appropriate thread for my question.
So here we go.
How does one avoid the 32 WU qouta limit ?
My machines are constantly running out of work, so i modified the boinc source to report more cpus(twice as much) than the computer actually has.
Is that ok or are there any reasons why i shouldn't do this ?
1st let me say to you that I’m sorry for what happen in the Seti Forum. Even if I had never crunched for Seti, my team informed us about the mess, then I went there to read the story. Seems like all kids are meeting on Seti's forum.
I'd like to thank you for your contribution to BOINC, and I'm very glad to see you here. You will, I'm sure, appreciate the great work Akosft done in optimizing Einstein.
What lead us to the 32 WU quota limit. I'm not as specialist as you are, but tweaking the number of CPU's should have work. It's what DocMaboul is doing in BOINCStudio, and it's working fine for everybody using it. Maybe he's doing it a different way than you're doing. Someone else should be able to tell you more.
If you want to read about BOINCStudio it's here: http://einsteinathome.org/node/191137
As the main forum of BOINCStudio is in French, if you have any questions, post it in this thread, we will translate it to the french forum and to BOINCStudio author.
Welcome Crunch3r! Sounds like you've figured it out by doubling the cpu count. That's what I had to do using BoincStudio. Setting the cache to 3 days seems to work out nicely, too. I actually survived the last outage with a few hours work left.
seems as if this is the appropriate thread for my question.
So here we go.
How does one avoid the 32 WU qouta limit ?
My machines are constantly running out of work, so i modified the boinc source to report more cpus(twice as much) than the computer actually has.
Is that ok or are there any reasons why i shouldn't do this ?
Hi Crunch3r,
Wellcome on E@H and thank you for your great job on Seti.
I'm the author of number of cpu faking used in BoincStudio. It's very easy to do. Just modify the number of cpu in int HOST_INFO::write(MIOFILE& out) function (hostinfo.c).
You have a new chalenge on einstein: try to do better than Akosf...
Edit: try to use BoincStudio. It's a very usefull tool for someone like you who have a large number of boxes. You can deploy a new optimized client in one clic.
No problem with the credit claim for a S@H enhanced WU about the same as with C3r 5.2.13, but I´ve heard that if you use Trux you skuold diaable it for S@H enhanced since tx36 dosen´t support that onte just keeping it for S@H standard if it´s possible.
Yes, you can read about it in some postings. But it's not necessary. The calibration does not have an effect on the credit claims of seti_enhanced.
Maybe the people got confused about the messages they got in their message tab. Of course, Trux's client still tries to calibrate the credits and cpu time and benchmark results are adjusted. But the credit system of seti_enhanced does not care about these numbers. seti_enhanced only looks at the fpops counting. These numbers are reported unchanged. Therefore it doesn't matter if calibration is turned on or off.
During the transition of seti to seti_enhanced I kept calibration turned on until it got useless. My seti_enhanced results never claimed different credits than other results crunched with 5.2.6+ clients.
Sorry, but can we keep posts on TOPIC. This is an Einstein@Home forum.
Idefix, you will find that the use of a trux calibrating BOINC client and a Akosf E@H client is ideal. What happens with SETI is really not relevant to this BB.
Sorry, but can we keep posts on TOPIC. This is an Einstein@Home forum.
Idefix, you will find that the use of a trux calibrating BOINC client and a Akosf E@H client is ideal. What happens with SETI is really not relevant to this BB.
FD.
I think that was uncalled for. He was still discussing an optimized client and that is what this thread is about
RE: RE: Just installed
)
I am sure the Crunch3r client supports the new credit system used by Seti Enhanced.
No problem with the credit
)
No problem with the credit claim for a S@H enhanced WU about the same as with C3r 5.2.13, but I´ve heard that if you use Trux you skuold diaable it for S@H enhanced since tx36 dosen´t support that onte just keeping it for S@H standard if it´s possible.
RE: I`m getteing a little
)
That is an amazing jump up ! *smile* I hope that happens to me as well.
Jonathan
Hello, seems as if this is
)
Hello,
seems as if this is the appropriate thread for my question.
So here we go.
How does one avoid the 32 WU qouta limit ?
My machines are constantly running out of work, so i modified the boinc source to report more cpus(twice as much) than the computer actually has.
Is that ok or are there any reasons why i shouldn't do this ?
Hi Crunch3r. 1st let me
)
Hi Crunch3r.
1st let me say to you that I’m sorry for what happen in the Seti Forum. Even if I had never crunched for Seti, my team informed us about the mess, then I went there to read the story. Seems like all kids are meeting on Seti's forum.
I'd like to thank you for your contribution to BOINC, and I'm very glad to see you here. You will, I'm sure, appreciate the great work Akosft done in optimizing Einstein.
What lead us to the 32 WU quota limit. I'm not as specialist as you are, but tweaking the number of CPU's should have work. It's what DocMaboul is doing in BOINCStudio, and it's working fine for everybody using it. Maybe he's doing it a different way than you're doing. Someone else should be able to tell you more.
If you want to read about BOINCStudio it's here:
http://einsteinathome.org/node/191137
As the main forum of BOINCStudio is in French, if you have any questions, post it in this thread, we will translate it to the french forum and to BOINCStudio author.
You might find informations in this thread too:
http://einsteinathome.org/node/191237
Sorry not to answer directly to your questions, and to give you a lot to read in return.
Once again, welcome to you.
Alex,
God created a few good looking guys.. and for the rest he put hairs on top..
Welcome Crunch3r! Sounds like
)
Welcome Crunch3r! Sounds like you've figured it out by doubling the cpu count. That's what I had to do using BoincStudio. Setting the cache to 3 days seems to work out nicely, too. I actually survived the last outage with a few hours work left.
RE: Hello, seems as if
)
Hi Crunch3r,
Wellcome on E@H and thank you for your great job on Seti.
I'm the author of number of cpu faking used in BoincStudio. It's very easy to do. Just modify the number of cpu in int HOST_INFO::write(MIOFILE& out) function (hostinfo.c).
You have a new chalenge on einstein: try to do better than Akosf...
Edit: try to use BoincStudio. It's a very usefull tool for someone like you who have a large number of boxes. You can deploy a new optimized client in one clic.
Hi, RE: No problem with
)
Hi,
Yes, you can read about it in some postings. But it's not necessary. The calibration does not have an effect on the credit claims of seti_enhanced.
Maybe the people got confused about the messages they got in their message tab. Of course, Trux's client still tries to calibrate the credits and cpu time and benchmark results are adjusted. But the credit system of seti_enhanced does not care about these numbers. seti_enhanced only looks at the fpops counting. These numbers are reported unchanged. Therefore it doesn't matter if calibration is turned on or off.
During the transition of seti to seti_enhanced I kept calibration turned on until it got useless. My seti_enhanced results never claimed different credits than other results crunched with 5.2.6+ clients.
Regards,
Carsten
Sorry, but can we keep posts
)
Sorry, but can we keep posts on TOPIC. This is an Einstein@Home forum.
Idefix, you will find that the use of a trux calibrating BOINC client and a Akosf E@H client is ideal. What happens with SETI is really not relevant to this BB.
FD.
RE: Sorry, but can we keep
)
I think that was uncalled for. He was still discussing an optimized client and that is what this thread is about
98SE XP2500+ @ 2.1 GHz Boinc v5.8.8
