May 27, 2009 00:23
On Friday we had a major server meltdown at work (source code repository), only to discover that the reigning sys admin had gone on vacation without communicating the latest password change, since they were still in the middle thereof. Travel to the co-location of the server and a hard reset were needed to get the developer team unsnagged.
This morning, we fixed the problem with a simple procedure change: Now the updated password sheets are distributed before the changes across all of our servers are actually implemented. This way, if the new one does not work on a particular machine, it is because it has not been rolled out there yet, so everyone can simply go back to the one they have already memorized and still get in.
Case solved.
security,
computer