Info about the disruption of IT services on October 24

On Tuesday, October 24, many services of the IT Services were (partially) unavailable from 11 a.m. to 4. p.m.. With this report, we would like to explain the reasons for the problem.

Almost all services of the IT Services (among others ILIAS, web server, GROUPS/BSCW and wiki server) are connected to a central storage system. The services thus depend on the smooth functioning of the central storage system, which makes it a highly reliable high-performance system with high redundance regarding data retention. Thus, the disruption on Tuesday was not due to a failure of any technical component of the system. We also had not planned any maintenance work. We would have announced that in adavance, but we had not planned for that at the beginning of the semester anyway. In fact, there was a much smaller technical issue which prevented the move of another service to the central storage system. That is why we contacted the provider of the system and discussed poosible ways of solving the issue. The provider recommended we change certain configurations in the system and also the running of a process for the system analysis. This was supposed to have a minor impact on the performance of the system. Unfortunately, that was not the case.

After the storage team had heard about the first disruptions, the analysis process was stopped. However, this did not help the situation. Only after intense troubleshooting together with the provider was it possible to detect the bug, which was caused by certain constellations in the system configuration and thus was not to be expected. Due to the bug, requests to the storage system could only be answered with a big delay. By resetting the configuration, the normal system behaviour was restored.

The IT Services undertands that the disruption was a big inconvenience for many users of the services. We apologise for that.

Your IT Services Team