3550 as ntp server - client still unsynchronized

[sthon-dbsys]

Hi togehter,

using an 3550 with IOS 12.1(11)EA1 as ntp server (cat1).

The client, a 3550 with IOS 12.1(11)EA1 (cat2) is the most time unsynchronized.

cat1 is always synchronized.

in cat2 you see in the log timestamp, that sometimes you have a "." at the beginning of the line and sometimes not.

The line with the "." means, that ntp is unsynchronized, a line without "." means ntp is synchronized.

Other devices, e.g a CSS get´s the correct time via sntp from cat1.

Anyone has seen this strange behaviour before ?

thanks in advance

sascha

[drolemc]

The point at the beginning of the timestamps indicates that the clock was synchronized before but lost contact with the server. I had a look at bugs related to ntp (for 12.1(11)) but I could not find any relevant to this problem. That leaves the possibility of connectivity issues between the device that is loosing synch and the server.

[sthon-dbsys]

I also found no bugs regarding this issue.

Some of these switches are directly connected, and I see no problem sof the connection, so I can eliminate connection problems.