Status as of Wednesday, September 20, 2017

[Steve Sokolowski]

Good morning!

• Customers have reported that there is a problem with hashrate recording for some users. It's unlikely that it's caused by earnings themselves, because if there were missing earnings, our profit would have been extremely large, and it was not. It is also possible, and perhaps likely, that the expected 40% loss in profits due to the exchanges being offline is combined with a display error such that the right amount was earned for what appears to be the wrong reasons. Chris will investigate in about 8 hours to determine if there is an issue and why these particular users are affected. Should a database issue that resulted in lost earnings be identified, Chris will make sure that corrections are issued.
• Many exchanges are either performing "maintenance" or are offline, cutting into profits. All pools are likely to have low profits until this issue is resolved by the exchanges, since many coins are only listed on one exchange. For those coins that are listed on more than one exchange, the prices obtainable are going to be lower because we can't select the exchange with the highest price. When the exchanges have fixed their systems, we will be able to resume normal coin selection.
• We expect the lawyer to get back to us within 7 days, and then we probably will need another 7 or so to evaluate changes (if any) based on his advice, and a few more days before we could make an announcement about what we decided to do. We will also need to adjust our numbers to account for the opportunity cost caused by the growth of competitors during this evaluation period (if any) and the number of lost customers who become frustrated with the site who will never return no matter how much we fix in the future. We spent over $10,000 to make sure that any possible legal issues will be identified, and we apologize for the continued delays while we evaluate the project's expansion or contraction prospects.
• We think that the "miners getting stuck on coins" issue has been resolved. That problem was caused by code that was added to reduce CPU usage in coin assignment, and a release was issued two days ago to fix it. The "pool dead" issues that some miners are experiencing are caused by DDoS protection. Those issues will not be resolved for at least one month, because the solution is too expensive to purchase unless we decide on expansion. Users affected by the "pool dead" issue should submit a support ticket to make sure that the problem isn't something easy, but if it isn't, then waiting will probably be your best bet.
• It's becoming obvious to me that the support ticket system needs to be replaced. There are issues with E-Mails containing special characters causing delays in tickets being received, broken filters, and so on. Many of the replacement systems are non-free and therefore we will wait for the decision before moving forward with this fix. In the meantime, the existing system should be acceptable, just not ideal.
• Also, please note that if you are submitting support tickets but feel that you aren't getting a reply, check that your ISP is not blocking messages from prohashing.com. Microsoft's spam filters, for example, block messages from the ticket system, so you either need to choose a different E-Mail provider, contact Microsoft to ask them to remove the block, or use the ticket system itself.
• At last count, there were something like 28 private messages backlogged. I wanted to remind everyone that private messages are not an official support channel, and they are answered on a "best effort" basis only. The problem with these messages is that because they are private, they are not always available to the most qualified person to answer them. Guaranteed responses answered by the most qualified person are only available by submitting a support ticket at contact@prohashing.com or calling the number on the contact page.
• There are a few issues that we will need to fix this weekend or next. One involves large numbers of miners connecting and disconnecting from the same IP address too frequently. There isn't really a reason why it should be necessary for one IP address to issue 10,000 connection requests per minute, and this issue can be handled using fail2ban with a very high limit. A secondary issue is that it's possible for a large number of static coin miners to mine the same coin, such that everyone's revenue is decreased due to ridiculous competition for blocks. Our main discussion point here will be figuring out which miners to disconnect when there are too many static miners for a coin and how to make it clear to them why they were disconnected (because there is no way in the stratum protocol to indicate that.)

[SpacedCowboy]

On the last point. I hope my miners aren't causing the issue (I have a bunch of L3+'s) but if so, my miners are co-located, so I don't get to see anything on their screen anyway. If you are going to do something, perhaps make the website the mechanism for giving out information... Maybe re-order problematic miners at the top of the list of "Live worker status" and flash them red/normal if they're being banned for some reason, along with a message. In other words, something blatantly obvious and in your face when you check to see your profits.

On a slightly different note - as I said, I have a bunch of L3+'s, and I have litecoinpool set up as a backup to this pool, so when the miner can't connect, it'll fail-over to the secondary (and then tertiary) pools. I notice that no-matter-what, even if this pool is "up" all night, I'm seeing LTC being mined over at litecoinpool every night by 4 or 5 of my L3+'s (different ones each time). Whenever it gets above 0.1LTC I clear it out, so I can keep track of the failovers, and I'm doing that pretty much every day these days. Here's a dump of the last 24 hours:



(By the way, the figures don't look very high, but that's because they're averaged over 24 hours. If you take 11,461 kHash/sec, then multiply by 86,400 (seconds in a day) then divide by 500,000 (kHash/sec for an L3+) you get ~1980 seconds, so that miner couldn't connect for roughly half an hour in the last 24 hours)

This is all FYI - maybe this is "as expected" for a pool, I'm not sure, but in a perfect world, I'm assuming there'd be no connectivity to litecoinpool as a backup...

[Steve Sokolowski]

On the last point. I hope my miners aren't causing the issue (I have a bunch of L3+'s) but if so, my miners are co-located, so I don't get to see anything on their screen anyway. If you are going to do something, perhaps make the website the mechanism for giving out information... Maybe re-order problematic miners at the top of the list of "Live worker status" and flash them red/normal if they're being banned for some reason, along with a message. In other words, something blatantly obvious and in your face when you check to see your profits.

On a slightly different note - as I said, I have a bunch of L3+'s, and I have litecoinpool set up as a backup to this pool, so when the miner can't connect, it'll fail-over to the secondary (and then tertiary) pools. I notice that no-matter-what, even if this pool is "up" all night, I'm seeing LTC being mined over at litecoinpool every night by 4 or 5 of my L3+'s (different ones each time). Whenever it gets above 0.1LTC I clear it out, so I can keep track of the failovers, and I'm doing that pretty much every day these days. Here's a dump of the last 24 hours:



(By the way, the figures don't look very high, but that's because they're averaged over 24 hours. If you take 11,461 kHash/sec, then multiply by 86,400 (seconds in a day) then divide by 500,000 (kHash/sec for an L3+) you get ~1980 seconds, so that miner couldn't connect for roughly half an hour in the last 24 hours)

This is all FYI - maybe this is "as expected" for a pool, I'm not sure, but in a perfect world, I'm assuming there'd be no connectivity to litecoinpool as a backup...

There should not be any failovers. The problem is caused by the VPNs and DDoS protection, and will be fixed by the new Internet connection if we decide on expansion.

If we decide on expansion, we should be able to have this problem fixed in a month or so.

[biscayne]

Since 3PM CET this afternoon we get stable hashrates again, thanks Steve and Chris!