I remember an issue with the Oracle DB client libraries for Linux (RHEL 2 or 3 I think) where if the system had something like a 160 day uptime any software usi

I remember an issue with the Oracle DB client libraries for Linux (RHEL 2 or 3 I think) where if the system had something like a 160 day uptime any software using the library to connect to the DB would just hang. I saw the effects with strace where it would just get stuck in a loop doing some time related syscall. We managed to get an unofficial patch from Oracle that I could wedge into our servers, since reboots were under strict change control. I remember this issue came up when I was being interviewed for a position at the Guardian a couple of months later and they seemed amused by what I had to do to fix it since they also encountered the same issue but fixed it by making sure they reboot more than twice a year.