Hi,
my HP 2900 switches has a time-synchronization issue. Look at the log:
I 02/27/09 22:03:55 mgr: SME SSH from 192.168.0.29 - MANAGER Mode
I 02/27/09 22:04:18 mgr: SME SSH from 192.168.0.29 - MANAGER Mode
I 02/27/09 22:06:23 mgr: SME SSH from 192.168.0.29 - MANAGER Mode
I 01/01/90 01:00:00 SNTP: updated time by -604619044 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 02/28/09 00:56:03 SNTP: updated time by 604626243 seconds
I 01/01/90 01:00:00 SNTP: updated time by -604634163 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 01/01/90 01:00:00 SNTP: updated time by -720 seconds
I 02/28/09 03:56:03 SNTP: updated time by 604637043 seconds
Time/Date is reset to 01/01/90 01:00:00
It happens every night during the same time-window.
I know that server backups are running during the night, causing some network traffic, but not so much (i didn't get an alarm about high drop rate / overbandwidth).
NTP server is a Linux Box, and from its logs all looks fine. Other devices syncing with it does not experience the strange time-reset issue. It happens only on ALL my procurve switches (even on the 8212 Core!).
Switch sntp config is as follows:
timesync sntp
sntp unicast
sntp server priority 1 192.168.0.200