- Steve Sokolowski

- Archived posts: 4559
- Joined: Aug 27, 2014
System catching up
There was a problem overnight where the share inserters stopped inserting data into the database. As designed, the system stored the shares in memory until someone was able to address the problem. The shares are being inserted right now and should recover throughout the next few hours. During this time, the reported numbers will be all over the place, because they are processing old data.
No money was lost, and we will correct the charts later in the morning. Feel free to reply if you have any additional questions.
Someone asked why this seems to happen when releases occur. The short answer is that unlike, say, a game, where you can program it and test it on your computer before shipping it, there is no way to test the huge number of miners that connect to this system before release. We can only simulate "fake" miners, which will catch some bugs but not ones due to network conditions, or specific types of miners, or attacks like the one on the Bitconnectcoin network.
Fortunately, I'd like to hope that we've gotten better at it. This release was more complex than the x11 release in April and resulted in just this one issue, which I've already identified the cause of and am looking into now. I don't think there are a large number of bugs that still remain to be found, as this same issue caused two crashes and there were 8 hours between.
No money was lost, and we will correct the charts later in the morning. Feel free to reply if you have any additional questions.
Someone asked why this seems to happen when releases occur. The short answer is that unlike, say, a game, where you can program it and test it on your computer before shipping it, there is no way to test the huge number of miners that connect to this system before release. We can only simulate "fake" miners, which will catch some bugs but not ones due to network conditions, or specific types of miners, or attacks like the one on the Bitconnectcoin network.
Fortunately, I'd like to hope that we've gotten better at it. This release was more complex than the x11 release in April and resulted in just this one issue, which I've already identified the cause of and am looking into now. I don't think there are a large number of bugs that still remain to be found, as this same issue caused two crashes and there were 8 hours between.
Last edited Thu Dec 07, 2017 01:00 PM UTC by Steve Sokolowski; edited 1 time(s).
