Dear Players,

as you all noticed, obviously, we have been experiencing major technical issues lately. In this post, we'd like to shed some light upon that.

But firstly, we'd like to thank you for all the supportive voices and offers of help, especially those to whom we couldn't respond personally due to the lack of time. We really appreciate all the good words, they've been the greatest motivator for us in this struggle. Secondly, we would like to take this opportunity to assure all players who recently joined Tibiantis that the last few days are highly unrepresentative of how Tibiantis usually operates. In fact, this is the first such situation since the server's launch in 2020. Back then, we also experienced an issue that forced us to restart the server twice. A lot of people were put off, understandably, but we managed to handle it aptly, and since then until just recently, we have not had ANY so-called "rollbacks" or other major issues. We'd like to emphasize this fact also to our long-time players, and ask them not to lose their trust in us. Technical issues affect all games and every companies from time to time. While every effort is made to avoid them, please realise that this is not always possible. Despite this being quite a large issue, we still think that once every 3 years is a relatively low and favourable rate. ;)



In the recent weeks we were forced to perform a rollback four times: Dec 13th (ca. 2 hours), Jan 3rd (ca. 2 h), Jan 7th (8 h), and Jan 8th (6 h). We are fully aware of the amount of disappointment and frustration this brought about, and, again, we apologise for all the inconvenience and emotional distress. What is important here to note is that these four issues were NOT one and the same problem that's been around for over a month, nor were they caused by one. The first two cases were the result of the hosting company simply shutting down one of our machines for their own maintenance without notice. Normally, such situations were preceded with a notification about the scheduled works, but this was not the case here. Therefore, to prevent similar cases in the future, we began to monitor their work schedule continuously instead of just relying on the notifications. This concludes the story of the first two rollbacks. There is really nothing more to that, and any further problems had nothing to do with it - just an unfortunate coincedence that they happened one after the other.

The issues from 7th and 8th January, however, did have one common cause, and it was the first time it occured. We managed to backtrace and reproduce the direct reason for the server going zombie, which involves a bug in some old external libraries (that we cannot easily update without probably breaking the compatibility elsewhere). We cannot yet pinpoint the initial trigger for the whole chain reaction. Because it has never come up before and the game server's code hasn't been changed in a long time, it must be caused by a very rare edge scenario that has just never occurred in the past 3 years. But to cut your losses and avoid another rollback in case this issue be triggered again, we implemented and successfully (so far) tested a procedure of an emergency save. Should such a situation recur, the game server will be brought to a halt, all players will be logged out safely, and all the characters and the map will be saved just like during a regular save. The only consequence will be a short reboot WITHOUT a rollback. In other words, you will not lose your progress - not even one lootbag because the map will be saved without the refresh. On that account, the game server will be back online today (Jan 10th) at 10 p.m. CET.

Regrettably, because we didn't have such a solution before, the game will now have to pick up from the last stable save, i.e. from 8th January, 09:00 a.m. CET. This refers to the in-game changes only - all the changes made after that time by the website (including all transactions) remain unaffected. House auctions have been extended so that the remaining time corresponds to that which was left before. All premium accounts, as well as accounts that were active that day, also received extra premium days. As a side note, we mentioned previously a possible change of the hosting machine, but we're dropping this idea for now. Thorough diagnostics revealed no hardware issues, and eventually we were able to reproduce the same problem on a testing machine, so changing the host now would be pointless.

Your Tibiantis Online Team