Yesterday at 21:40 GMT we experienced an outage related to our Redis systems, which caused Miki to shut down for a 2 hour timespan, and a loss of data. Because of this cache failure, we have lost specific temporary, but important data, but most impactful: Streak data for circa 50% of the active streaks.
We understand that this kind of data loss is unacceptable; especially to you those of you that almost spent a full year into dedicating yourselves to Miki and her daily rewards.
Can we fix it?
Is probably what is most important to you, and sadly; with our current tech, we can't. We don't have either the manpower to provide all of you with a solution and we don't have the tech to roll back a temporary storage engine.
I would like to apologise for this, and hopefully we can meet in a midway point for this.
This has happened before, why did it happen again?
There's a lot of metrics that play into one of these issues; most of the time you, the user won't notice that Miki is running into issues somewhere due to our error-prone architecture, but this is the only part that we can't easily change.
We will however, promise to put priority into this issue and will not happen again.
I understand it's hard to trust this, yet again; but I can only prove it to you in actions.
P.S. if you ever need to contact me for anything miki-related. My email is open and you're allowed to email me whenever you want. [email protected]