We regret that there was a problem with ChemNet wifi between 8am on 9th March and 10.30 on 10th March which caused connections to fail to be assigned IP addresses, and so people would have found they had no network access despite their machine reporting being connected. This was fixed at 10.30 on 10th March and things are now running normally.
The problem began when we restarted the DHCP server (the software that allocates addresses) to update the list of DNS servers it sends out with the IP addresses. This is a routine process, and we’d done it several times recently as we have been updating our DNS servers. None of the previous restarts had issues. This time the DHCP software appeared to restart as normal, without any error message, but for some reason did not then allocate any IPs until a second restart was done. Unfortunately none of the IT staff were in the building on the 9th March so we didn’t become aware of the problem until someone reported it to us on the 10th.
We are looking into providing better monitoring of that DHCP server to detect failures automatically in future.