The server that cbhistory runs on went down this afternoon due to a power failure. It had been up for over 205 days.
Power was restored within a minute and the machine rebooted without incident - except that it's NTP server was down due to the massive power failure now affecting NYC and much of the northeast US. I guess the hardware clock drifted quite a bit over the past 7 months. With no NTP server to reset the system time, it came up with the time set about 15 hours off. As a result, as soon as cbhistory would write new records to the database, it purged them back out again thinking they were over an hour old.
I just pointed xntpd at a server in Pennsylvania and the time is now correctly set. Cbhistory should be running properly again. ++bart for pointing out that it was down.
|90% of every Perl application is already written. ⇒|