Quote from: petabytecoin on September 19, 2012, 03:59:19 PM

Quote from: philips on September 19, 2012, 09:41:36 AM

Still having occasionally comunication failures and plenty of Pool not providing work fast enough. The very annoying thing is that is never bad enough for cgminer to consider the backup pool but bad enough for hashrate to drop (failover strategies).

Load balance strategy seems to always favour the backup pool even when CoinLab is doing OK, so no.
Balance strategy works, but with a hashrate often significantly reduced (70-50%), so no again.
The site is also sometimes unreachable.
NO, I do not have any net/software/hardware problems.

Any ETA on EU servers?

I have had same experience with comm failures, pool not providing work fast enough, etc. What is the bottleneck for this, bandwidth? Server power on their side? I'm not that far geographically from them so that's not it. I also don't get these errors on other pools that are much farther away.

Sorry about the growing pains we've been having with our pool. There are two sources for these communication failures.

1) We had capacity issues with our servers - causing them to slow down.
2) When we update our server software, we drop connections temporarily.

As we've gotten more demand for our pool, we've had to add additional front-end servers, and today we switched to high-cpu instances on Amazon. We are trying to get ahead of the curve - as of today I think we've doubled our capacity, which should reduce latency and allow more miners to connect w/o overloading our servers again.

I want to assure you that even when we have intermittent communication failures, you should not be losing any shares. We have a lenient 90-second share acceptance policy; so as long as your miner is able to re-establish a connection to our pool within 90 seconds from the time we gave you the work, you will still get credit for it (even if it is stale).

Thanks again for your feedback, patience and helping us improve our pool!

Mike Koss
CTO, CoinLab