Topic: 100 % cpu bug on Slush's pool (Read 1757 times)
cgminer does seem to be fixing the cpu bug.
cgminer. (yes that is a period)
Wouldn't the simplest solution be to drop GUI Miner and use a real one?
Do you have any suggestions on what a real one might be. From my understanding guiminer is just a front end for the real miners. And No I'm not going to be switching to linux anytime soon since I use my PCs for more than just mining.
I am mining on slush with a 3 cards rig + a 2 cards rig with no issues using aoclbf (so phoenix) with no issues at all in 2 months.
Wouldn't the simplest solution be to drop GUI Miner and use a real one?
(repost from pool thread)
This is very interesting, but some other people come from other pools to my pool exactly because they have no 100% cpu bug with me :-). I mean - there's probably some strange issue inside guiminer which can be triggered by something like faster/slower responses etc (I mean some strange race condition or so). I didn't changed mining part since March 16 and I even cannot imagine how can I affect this bug by any change inside the pool...
This is very interesting, but some other people come from other pools to my pool exactly because they have no 100% cpu bug with me :-). I mean - there's probably some strange issue inside guiminer which can be triggered by something like faster/slower responses etc (I mean some strange race condition or so). I didn't changed mining part since March 16 and I even cannot imagine how can I affect this bug by any change inside the pool...
It started sometime overnight between the 16th and 17th of March, I know this because I built the 8120 System on the friday evening and it was working fine with no cpu bug when I went to bed, in the morning all my Miners were doing the kilohash thing till I checked off all the cores.
Im in the GMT -6 time zone.
Im in the GMT -6 time zone.
No CPU bug on Eligius or Deepbit either
BTC Guild is just that cool!
That's a very interesting find...I can't imagine what's going on that would make the bug appear based on the specific pool you're pointed at. Have you tested any others?
That's a very interesting find...I can't imagine what's going on that would make the bug appear based on the specific pool you're pointed at. Have you tested any others?
Heres some screen shots of this in action:
Mining on slush with all cpus selected works ok uses 100% of a core
http://imageshack.us/photo/my-images/19/btc1.jpg/
With all cpus unselected kilohash range mining on my 6870
http://imageshack.us/photo/my-images/836/btc2.jpg/
Mining on Btcguild all cpus selected almost no cpu usage
http://imageshack.us/photo/my-images/171/btc3k.jpg/
Mining on Btcguild no cpus selected almost no cpu usage no problems
http://imageshack.us/photo/my-images/580/btc4.jpg/
Mining on slush with all cpus selected works ok uses 100% of a core
http://imageshack.us/photo/my-images/19/btc1.jpg/
With all cpus unselected kilohash range mining on my 6870
http://imageshack.us/photo/my-images/836/btc2.jpg/
Mining on Btcguild all cpus selected almost no cpu usage
http://imageshack.us/photo/my-images/171/btc3k.jpg/
Mining on Btcguild no cpus selected almost no cpu usage no problems
http://imageshack.us/photo/my-images/580/btc4.jpg/
no I was not cpu mining but when I am connected to slush if I don't check off all the cpu affinity checkboxes then it drops into the kh/s range mining on my GPUs on my 8120 I check off all 8 because it seems to randomly choose a core to use and if they are not all checked it drops down to the kilohash range if it doesn't have the one it picks available
in the kilohash range indicates you were cpu mining. your settings must be wrong, double check you're mining on gpu and not cpu.
No the 100% bug has been affecting me for a couple weeks on all my machines they were all not having the problem then one morning I checked them and they were hashing in the kh/s range until I enabled more core afinity on them.
I just noticed that it doesn't affect me when I connect to btc guild because of the problems with slush's pool tonight before that I hadn't tried another pool.
I just noticed that it doesn't affect me when I connect to btc guild because of the problems with slush's pool tonight before that I hadn't tried another pool.
I'm guessing that it's just continually trying to reconnect. See the slush pool thread, but the pool is down.
Hows this for weird behavior:
Using Guiminer on Win 7 64 on 3 different machines, using 11.12 and sdk 2.5 and opencl flags in gui miner are -v -f 90 -w 128
When I mine on Slush's pool I get the 100% cpu bug uses one core full time on my fx8120 and the same on my dual core athalons for the other 2 machines
If I switch to btcGuild no cpu bug same settings.
Any ideas?
Using Guiminer on Win 7 64 on 3 different machines, using 11.12 and sdk 2.5 and opencl flags in gui miner are -v -f 90 -w 128
When I mine on Slush's pool I get the 100% cpu bug uses one core full time on my fx8120 and the same on my dual core athalons for the other 2 machines
If I switch to btcGuild no cpu bug same settings.
Any ideas?
Jump to: