Tomorrow Tue Nov 3, starting from 0730 we will upgrade our eGroupware service to version EPL 9.1. Expected downtime: about one hour.

Today at 09:50 a crucial server died in our HIT server room. It took us about 20 min to move the affected services to other machines, during which time most of our machines weren't usable. We're sorry for the inconvenience. We're working hard to get rid of this fault-prone hardware.

On Wed Sep 23 around 01:00 a major power outage hit Höngg and Affoltern. Our servers on the Hönggerberg campus switched to battery power, but since power was gone for more than 3 h, the batteries eventually drained and all servers switched off. Around 07:15 we began rebooting our infrastructure servers, but it wasn't until noon that the most important services (home, web, mail) were back up. Please bear with us while we still iron out the last remaining issues.

Tomorrow Tue September 8, starting from 0730, we will upgrade our Roundcube installation. Expected downtime: about 15 min.

Update Tue 0734: upgrade done.

In case you've been wondering about the slightly dubious announcements of the past few days: on Friday (2009/08/14) a local privilege escalation in all Linux kernels of the last 7 years was published, together with an exploit. Unfortunately no patched kernels were available by Friday late afternoon, which put us into an awkward position. Generally it is not our policy to be sneaky about security issues, but in this case we really did not want to attract malicious script kiddies. That's why we decided to keep our announcements somewhat vague. By now the worst seems to be over and all machines have been rebooted with patched kernels.

We apologize for any confusion or service degradation this episode may have caused on your side.

Due to a critical condition on our systems we were forced to shut down our backup servers. This means there will be

No backup during this weekend.

Please be aware of this temporary limitation when producing large or important amounts of data. We apologize.

Today at 07:04 a software upgrade of a core router performed by Informatikdienste caused an avalanche of errors in our server cluster, leading to an outage of printing and some other non-critical services. It took us till around 08:00 to get the systems back to normal. We apologize for any inconvenience.

Today at 16:41 a process on our terminal server plimpy freaked out and consumed all memory + swap, causing a complete halt of all other processes. We had to reboot the machine. We're currently looking into ways to prevent this problem from appearing again in the future. Sorry for your inconvenience.

Today around 0715 we experienced a series of power outages in our water-cooled racks in HIT, leading to a shutdown of our most crucial servers. It took us almost two hours to get the systems up and running again, during which time most of our services were not available.

We apologize for any inconvenience.

Today around 14:20 the D-PHYS network started acting up. After about 20 minutes the problem could be tracked down to a machine running amok. Unfortunately, the terminal server plimpy did not survive the downtime and had to be rebooted at 16:10. We apologize for any inconvenience.