May 15
Scheduled Maintenance: May 17th
Next Monday, between 1 am and 2 am CEST (what time is that for me?), some updates to our servers will be made. Therefore, mite will be unavailable for a very brief period of time. We expect the interruption to last for no longer than 10 to 15 minutes.
Maintenance will include updating the kernels, i.e. the heart of the server systems, as well as some improvements to the hardware, i.e. server rack restructurings. This maintenance takes place to reduce the possibility of future downtimes by tackling the root of past problems. We ask for your understanding.
April 29
Today's service interruption
Between 1:22pm and 2:20pm CEST, mite was down for all users. We’re terribly sorry, please accept our apologies!
The reason for this downtime were problems in our data center: defect routers of the upstream provider caused an interruption in the external connection. Three minutes after this downtime started we began to update on this problem via Twitter. Within minutes, technicians started working on the hardware in the data center. Collaborating with our hosting partner SysEleven, we’ll keep looking into this problem to prevent similar problems in the future, this goes without saying. Of course, your data was totally safe throughout this downtime.
Again: we’re so sorry! This shouldn’t happen.
March 20
Scheduled Maintenance, take two: March 21th
Update, March 21th, 7:53am: Maintenance went as planned.
One of our brand-spanking new servers isn’t feeling quite at home yet. To prevent future problems up front, we’ll therefore move it to a new server blade. Unfortunately, this action requires another scheduled maintenance. Fortunately, this one requires a very short period of time only.
mite will be offline on Sunday, March 21th, between 7:30am and 7:50am CET. (What time is that for me?)
We’re so sorry for this second interruption! This decision didn’t come easy. But we’re convinced that this preventive step makes sense even if nothing serious happened yet. We ask for your understanding.
March 19
Today's downtime
At about 11am CET, mite became extremely slow; So slow, that the service went down between 11:24am and 11:35am. First of all: we are terribly sorry, this shouldn’t happen!
These problems were due to a disruption of the external internet connection of the data center where our servers are taken care of. Our own alarm systems did work, as well as the ones of our hosting partner SysEleven; we updated via Twitter. SysEleven managed to solve the problems of the IP uplink. All seems fine again. Nevertheless, you can definitely count on us for being extra-observing today.
Again: we’re sorry. Please excuse this interruption!
March 02
Scheduled Maintenance: March 7th
Update, March 7th, 5:18am: Everything went as planned. Good time tracking on the new servers, everyone! Just in case you stumble over a bug: please get in touch with as much details as possible.
Safe, secure and lightning-fast: that’s how mite behaves today and should behave in the future, no matter how fast the user base grows. To make this happen, Sebastian who’s taking care of the technical infrastructure here has been preparing mite for the next step: this weekend, we’re moving the application to a new server cluster. Therefore,
mite will be offline on Sunday, March 7th between 3am and 5am CET. (What time is that for me?)
We’ll update this blog post and keep you posted on Twitter in real time.
Ideally, you won’t notice a thing about the new infrastructure. All your data will stand at your service exactly as it is used to, this goes without saying. Nevertheless, experience shows that in production we might have to tweak the system here and there a little bit to optimize its responsiveness and stability – despite testing, testing, testing up front really thoroughly. Therefore, not only us, but also the team of SysEleven, our new hosting partner, will be extra beady-eyed. Promised. Now, let’s get moving!
June 09, 2009
Domain name problem affecting user subset
[Update: June 10th, 7:25 a.m.] The regular domain *.yo.lk which was erroneously suspended is now working again. You should be able to access your mite.account under its standard URL. As a matter of course, your data was safe at every moment.
It could take some time for DNS servers all over the world to reflect the unsuspension. If you cannot access your account by now, please continue accessing mite via the emergency domain. We are deeply sorry for these problems, even if we were not responsible directly for them. We know that you depend on mite. We will do everything possible to gain your trust again. Sorry.
—
Since this afternoon, some users are experiencing problems accessing their account. Our servers are up and running, but there is a problem with our DNS entry. If you happen to be affected, we want to apologize first of all!
We set up an alternative domain, under which you can access your account:
http://youraccountname.appmite.de
May 24, 2009
Downtime
Between 16:01 and 17:29 this Sunday afternoon, mite was down for all users due to a power interruption in our data center. We are terribly sorry for this interruption of the service! This shouldn’t happen. We’ll continue to look into the issue, together with the very capable people in Munich maintaining the infrastructure, to prevent something like this to happen again. Again: we are sorry for the inconvenience. Please stay with us.
May 22, 2009
Polished backend engine
Some minutes ago, we pushed an update of the underlying application framework. If you cannot notice anything working differently, we’ll put our thumbs up. Yes, the update speeds up mite a little bit and will help us maintaining the app, but it doesn’t feature changes on a functional level.
Nonetheless, if you happen to stumble upon a bug, please tell us so we can fix it right away. Send us an e-mail which describes the ‘corpus delicti’ as detailed as possible. And don’t forget to include your browser and operating system. Thanks!