Page 2 of 7

Re: Downtime this morning

Posted: Wed Jul 19, 2017 6:41 am
by biscayne
Pool down and sysadmins phone on mute again?

Re: Downtime this morning

Posted: Wed Jul 19, 2017 8:31 am
by Steve Sokolowski
biscayne wrote:Pool down and sysadmins phone on mute again?
No, this time the system wasn't on mute, so Chris got it online.

I hope to make some improvements on Saturday and Sunday this week.

Re: Downtime this morning

Posted: Wed Jul 19, 2017 8:43 am
by biscayne
Thanks, maybe he can use an alarm clock as it was about the same time as yesterday when it went down.

Re: Downtime this morning

Posted: Wed Jul 19, 2017 9:36 am
by jaybizz
Honestly, if the service as-is can't keep up with demand, you should think about blocking (or at least capping) rented hash from NH. Those of us with actual miners shouldn't be punished with lower quality service. At least until you're able to optimize things enough to handle the extra workload.

Re: Downtime this morning

Posted: Wed Jul 19, 2017 10:14 am
by AvPro
jaybizz wrote:Honestly, if the service as-is can't keep up with demand, you should think about blocking (or at least capping) rented hash from NH. Those of us with actual miners shouldn't be punished with lower quality service. At least until you're able to optimize things enough to handle the extra workload.
It sounds like it could be pure number of users / workers. Improvements to the Hashrate capacity of the pool were made earlier this month. I believe the pool can handle a good amount more than the hashrate that is presently connected

Re: Downtime this morning

Posted: Wed Jul 19, 2017 10:21 am
by Steve Sokolowski
AvPro wrote:
jaybizz wrote:Honestly, if the service as-is can't keep up with demand, you should think about blocking (or at least capping) rented hash from NH. Those of us with actual miners shouldn't be punished with lower quality service. At least until you're able to optimize things enough to handle the extra workload.
It sounds like it could be pure number of users / workers. Improvements to the Hashrate capacity of the pool were made earlier this month. I believe the pool can handle a good amount more than the hashrate that is presently connected
We deployed another fix last weekend to limit the number of shares per second that a miner can submit with a static difficulty. Miners who try to submit more shares are reassigned to a higher difficulty after five minutes.

Unfortunately, it would take just as much work to try to ban people and limit accounts as it would to simply start fixing the true issue. Our current thoughts are that we will raise fees soon to reduce the number of customers, and then lower them as capacity increases.

Re: Downtime this morning

Posted: Wed Jul 19, 2017 10:43 am
by vhmanu
i had hopes my short NH order would be fine. Had no rejections (started after order finished).
shareserver crashed a few minutes ago right at the part where those shares would have been processed :(
Whole order gone into nothingness :(

Re: Downtime this morning

Posted: Wed Jul 19, 2017 11:59 am
by jaybizz
Steve Sokolowski wrote:
AvPro wrote:
jaybizz wrote:Honestly, if the service as-is can't keep up with demand, you should think about blocking (or at least capping) rented hash from NH. Those of us with actual miners shouldn't be punished with lower quality service. At least until you're able to optimize things enough to handle the extra workload.
It sounds like it could be pure number of users / workers. Improvements to the Hashrate capacity of the pool were made earlier this month. I believe the pool can handle a good amount more than the hashrate that is presently connected
We deployed another fix last weekend to limit the number of shares per second that a miner can submit with a static difficulty. Miners who try to submit more shares are reassigned to a higher difficulty after five minutes.

Unfortunately, it would take just as much work to try to ban people and limit accounts as it would to simply start fixing the true issue. Our current thoughts are that we will raise fees soon to reduce the number of customers, and then lower them as capacity increases.
This particular issue, with high rejects/stales and whatnot, seems to have really cropped up in the past few days - has this "fix" been looked into as far as potentially causing the current issue?

And as far as raising fees to basically turn customers away... that is definitely not a good way to keep customers in the long run. There has to be a better solution than that.

Re: Downtime this morning

Posted: Wed Jul 19, 2017 12:58 pm
by CritterDog
A update on why the pool is dead again would be good.. If its Nicehash causing the problems they should block there ip address until the problem is figured out

Re: Downtime this morning

Posted: Wed Jul 19, 2017 2:11 pm
by glccom
I agree with blocking the IP's from NH, raising fees seems counter productive.