As a lot of you probably have experienced, we’ve seen a much lower level of performance than we’re generally used to. This came shortly after our migration to newer hardware within our vendor’s datacenter. We’ve done quite a bit to try and address the issues as the overall cause wasn’t immediately obvious. Partly because there was a certain expectation that came from being with the same provider, and partly due to several attacks against our services giving us a false perspective on the situation.
After some extensive monitoring after the attacks, and back-and-forth communication with our vendor, it has come to our attention that the new server is being thermally throttled under heavy loads causing a substantial decline in overall performance and general stability.
Obviously, this isn’t ideal. We’re currently doing everything in our power to try and make the situation a lot more tenable, but there’s sadly only so much we can do at this time. We are in active talks with our provider to get the situation rectified. But, that may take a bit of time.
I humbly ask for your patience and understanding as we do our best to improve the situation. In order to provide an overall better experience to our players, we will be temporarily reducing the population caps for our servers down to 40 players (donators will still be allowed to bypass this restriction like normal). Once the situation is rectified, this reduced cap will be reevaluated.
Again, I want to express sincere apologies for the overall poor performance as of late. Things will improve, it’s just going to take a little longer.
Update - May 12th, 2021
We’ve gotten the old server upgraded to be in-line with what we have now and have confirmed that the thermal throttling issue is no longer present on that hardware. We are currently awaiting additional network assignments from our datacenter before we can begin moving things to that environment.
Current projected date of completion: Friday, May 14th.
Update - May 14th, 2021
We have migrated to the new-new environment complete with better cooling and proper turbo-boosting. We’ll be monitoring things from here to see if we continue to get the expected performance.
Update - May 15th, 2021
I semi-secretly lifted the pop caps on Sage/Golden to see how things would perform. While it has been generally fine, we’ve observed some minor hiccups. Nothing overly concerning during the higher population tests. This has been somewhat counteracted by a DoS attack targeting one of our internal procs, we have implemented a temporary solution to help mitigate the problem and will seek a better, long-term, solution soon.
We are continuing to monitor the situation.
Update - May 19th, 2021
At this time, we’ve returned to a state of normalcy. The issue with abnormally high time dilation was resolved on the 17th, and we haven’t observed any further issues since. I’ve done several polls in-game and the respondents all seem to agree that we’re in a good spot once again.
This will likely be my last update on this issue. Thanks again to everyone for providing their support and for being patient whilst we worked on improving things.