Pages:
Author

Topic: [ Bonus PPS Pool - 130% PPS - OPEN ] - page 23. (Read 94008 times)

hero member
Activity: 504
Merit: 502
June 17, 2012, 07:39:19 PM

Are you running cgminer? Even weak cards would pick up again and you would notice HW: value increase from 0 for every lockup or other failures in the hardware.

I am running cgminer 2.4.2 on Xubuntu 12.4 (Catalyst 11.11 drivers with the included 2.5 SDK) on my rigs.

Unfortunately these locks are hard locks where one of two things happen:

1. I can log in under a root SSH session and see the cgminer PID is zombied to point even a Kill - Sig 9 will not kill it. Most of the time I can do a reboot command and gracefully reboot.

2. Cannot get a SSH session and have to hard reset, power-off & on.

Both rigs are using different MB and CPU and I separated them among two lightly used circuits in the house. I keep power supply demand at least 100 watts under rating.

For testing, I have reduced both rigs to 2 GPU and I have already found 1 card that locked up under this lighter condition.

Wow ok I guess even cgminer cant reset cards that are just purely into self destruction. Wink

Yeh seems you need to call the GPUlance to do some CPR on those cards.
legendary
Activity: 1064
Merit: 1000
June 17, 2012, 07:36:32 PM

Are you running cgminer? Even weak cards would pick up again and you would notice HW: value increase from 0 for every lockup or other failures in the hardware.

I am running cgminer 2.4.2 on Xubuntu 12.4 (Catalyst 11.11 drivers with the included 2.5 SDK) on my rigs.

Unfortunately these locks are hard locks where one of two things happen:

1. I can log in under a root SSH session and see the cgminer PID is zombied to point even a Kill - Sig 9 will not kill it. Most of the time I can do a reboot command and gracefully reboot.

2. Cannot get a SSH session and have to hard reset, power-off & on.

Both rigs are using different MB and CPU and I separated them among two lightly used circuits in the house. I keep power supply demand at least 100 watts under rating.

For testing, I have reduced both rigs to 2 GPU and I have already found 1 card that locked up under this lighter condition.
hero member
Activity: 504
Merit: 502
June 17, 2012, 07:36:09 PM
I had some weird shit happening... Seems fine now.. .Doubt it was due to the pool Smiley Things have been running smoothly for a while now.

Sorry that "weird shit" is coming from the toilet Wink , seriously its been running perfect without any hickups yet so definitely not the pool at fault this time Smiley
hero member
Activity: 1078
Merit: 502
June 17, 2012, 07:27:20 PM
I had some weird shit happening... Seems fine now.. .Doubt it was due to the pool Smiley Things have been running smoothly for a while now.
hero member
Activity: 504
Merit: 502
June 17, 2012, 07:14:42 PM
What kind of witchcraft you've used? Pool is now running super smooth for me with rejects averaging 1% and decreasing!  

I've also noticed drop in efficiency value computed by cgminer. I don't really know what that number really means, but the drop certainly indicates some change in the way the pool operates, am I right?

Well it means generally less getworks requested compared to submitted from pool. I have actually disabled all getwork queues to see if it improves stuff and it does by the look of things. As allways its way too early for this pool to say anything about improvements but lets enjoy it while its behaving. Im still fairly certain it will cough up bones at any time shrug.

Your are chasing bones, and I am chasing ghosts... Cheesy

Thought I had everything running smoothing, then after several hours the ghost of GPU lockup came knocking.

I suspect a weak card or two, I have already identified one, looks like about three more to check. Would be nice if the weak cards would just fail right away, but I guess it is good in the sense that if they take time to fail they will probably be fine with less aggressive settings...I hope.

Maybe I need the Ghost-busters........

Are you running cgminer? Even weak cards would pick up again and you would notice HW: value increase from 0 for every lockup or other failures in the hardware.
hero member
Activity: 1078
Merit: 502
June 17, 2012, 10:22:31 AM
Thanks for the heads up......
legendary
Activity: 1064
Merit: 1000
June 17, 2012, 07:08:25 PM
What kind of witchcraft you've used? Pool is now running super smooth for me with rejects averaging 1% and decreasing!  

I've also noticed drop in efficiency value computed by cgminer. I don't really know what that number really means, but the drop certainly indicates some change in the way the pool operates, am I right?

Well it means generally less getworks requested compared to submitted from pool. I have actually disabled all getwork queues to see if it improves stuff and it does by the look of things. As allways its way too early for this pool to say anything about improvements but lets enjoy it while its behaving. Im still fairly certain it will cough up bones at any time shrug.

Your are chasing bones, and I am chasing ghosts... Cheesy

Thought I had everything running smoothing, then after several hours the ghost of GPU lockup came knocking.

I suspect a weak card or two, I have already identified one, looks like about three more to check. Would be nice if the weak cards would just fail right away, but I guess it is good in the sense that if they take time to fail they will probably be fine with less aggressive settings...I hope.

Maybe I need the Ghost-busters........
hero member
Activity: 504
Merit: 502
June 17, 2012, 06:47:44 PM
Been mining with Clipse for a few weeks now, personally and with Cognitive.

Excellent uptime and very few rejects, thanks for the pool, and keep it up Smiley

I do appreciate this comment but sadly uptime(mainly down for few minutes at times due to changing things) and rejects(well this is still a huge mystery for me since its definitely unrelated to server load as I once thought it to be) have been horrible for last couple of weeks especially.

Eitherway thanks for the good words Smiley
hero member
Activity: 504
Merit: 502
June 17, 2012, 06:46:05 PM
What kind of witchcraft you've used? Pool is now running super smooth for me with rejects averaging 1% and decreasing!  

I've also noticed drop in efficiency value computed by cgminer. I don't really know what that number really means, but the drop certainly indicates some change in the way the pool operates, am I right?

Well it means generally less getworks requested compared to submitted from pool. I have actually disabled all getwork queues to see if it improves stuff and it does by the look of things. As allways its way too early for this pool to say anything about improvements but lets enjoy it while its behaving. Im still fairly certain it will cough up bones at any time shrug.
hero member
Activity: 1078
Merit: 502
June 17, 2012, 10:11:23 AM
That's good to hear...
newbie
Activity: 22
Merit: 0
June 17, 2012, 06:21:07 PM
What kind of witchcraft you've used? Pool is now running super smooth for me with rejects averaging 1% and decreasing! 

I've also noticed drop in efficiency value computed by cgminer. I don't really know what that number really means, but the drop certainly indicates some change in the way the pool operates, am I right?
legendary
Activity: 938
Merit: 1000
What's a GPU?
June 17, 2012, 05:49:35 PM
Been mining with Clipse for a few weeks now, personally and with Cognitive.

Excellent uptime and very few rejects, thanks for the pool, and keep it up Smiley
hero member
Activity: 504
Merit: 502
June 17, 2012, 02:27:05 PM
Thanks for the heads up......

Its mostly back miners should reconnect. Just checking over next few minutes how it runs else I revert back quickly to previous run.

Hey Clipse, I've had to switch to my backup fulltime. Not sure if you're still awake (and if I'm the only one) but I'm getting over 10% rejects in the last 4 hours along with regular periods where my GPUs are idling. Are you actively working on the pool, or is something broken?

Not doing any work for last couple of hours , its been running non-stop. It seems to me you may be the only one :/

EDIT: Can you try again, may have resolved it for you, a few IPs were getting rejected for some reason.
hero member
Activity: 1078
Merit: 502
June 17, 2012, 01:43:40 PM
Seems to be working on my end... I like bugging him on IRC Smiley 2.55% retards on my end
legendary
Activity: 1274
Merit: 1004
June 17, 2012, 01:42:06 PM
Thanks for the heads up......

Its mostly back miners should reconnect. Just checking over next few minutes how it runs else I revert back quickly to previous run.

Hey Clipse, I've had to switch to my backup fulltime. Not sure if you're still awake (and if I'm the only one) but I'm getting over 10% rejects in the last 4 hours along with regular periods where my GPUs are idling. Are you actively working on the pool, or is something broken?
hero member
Activity: 504
Merit: 502
June 17, 2012, 10:35:33 AM
Thanks for the heads up......

Its mostly back miners should reconnect. Just checking over next few minutes how it runs else I revert back quickly to previous run.
hero member
Activity: 1078
Merit: 502
June 16, 2012, 10:50:02 AM
HAHAHAHAHAHAHA... That was great....
hero member
Activity: 504
Merit: 502
June 17, 2012, 10:19:36 AM
Pool will be down for few minutes, again sorry for this but I rather take it down so you efficiently connect to backup pool while I sort out things otherwise you would continously reconnect which just sux.
sr. member
Activity: 366
Merit: 250
#RIP freemoney
June 17, 2012, 10:10:23 AM
Thanks for the updates. Was just about to add a lot of com failures too but you've explained that.

I hate satoshidice
hero member
Activity: 504
Merit: 502
June 17, 2012, 09:49:52 AM
DDOS NOTICE:

My datacentre have a couple of banking clients who current get hit by huge DDOS attack so they are busy resolving that so rest of clients(us) dont get affected. This should clear up soon, guess the DOSSing dildos have much more than just bitcoins on their mind. Its not our actual pool being hit.

EDIT: that was quick, network issues resolved.
Pages:
Jump to: