We've ran into a fair amount of problems this week which really does suck. We've been working and monitoring our servers for days at a time with little to no sleep. As we've been working on this large issue, we haven't been able to do much else so the changelog will explain what exactly is going on behind the scenes at the moment but first, here is the list of things we did change.
- Fixed Discord sync syncing the wrong account due to giving out duplicate codes
- Fixed a bunch of errors in the reward messages for crates
- Changed the prices of obsidian in the shop
- Removed double vote rewards
- Fixed you not getting the YouTuber tag from the prison points shop
- Added TP trapping to the rules
- Disabled you being able to create withers
- Pushed our gang islands and wars update
- Added a cooldown on crates transfer due to our logs being absoutely massive because people are spamming it every second
- We've recently hired another developer, Cameron, so if you see him about make sure you say hi to him!
We're aware of a few other issues such as generators grade upgrade breaking over a server reboot. These will be fixed soon.
With multiple accounts advertising their shops every 2 minutes, our chat has become stupidly spammy. You will only be able to put 1 message every 2 minutes if you are advertising your shop. This means you cannot send your messaged followed by a "." to bypass the chat filter.
We made a change to the way cactus works a few days a go. If you have no tier hoppers in your chunk, it will work as usual. If you have a tier 1 or 2 hopper in your chunk, you will not see the cactus growing, and you will just see it being added to your hopper. If the hopper gets full, it will grow and drop as usual. If you have a hopper with autosell, it will just be sold straight away. You will not see it enter the hopper. This allows us to keep running 950 slots on western at over 20 tps.
Recent lag issues and constant reboots (mainly Western)
For the last few months, we've been working on an issue where our servers are using a stupid amount of memory, which then causes our servers to freeze up when the garbage collection (freeing up memory) happens. These are why you get these spikes on Western. The TPS is running absoutely perfectly but because our CPU is working hard to free up memory, you get a 3-4 second spike every few minutes. We are working extremely hard to fix this, however, none of the others or any one else we have hired to look at it for us have any idea what is going on. It is completey unusual and we (or anyone else) has seen anyone like it before and there is no possible explanation for why it is happening. We are using the best software avaliable to us to try and figure out the issue but we are having to reboot Western a fair amount throughout the day to:
a) Push the changes we are making to see if it makes any difference