[Linux-cluster] weird time stamp in the message log file

[FM]

Our problem righ now is the speed of rsync + GFS.

Today looking in the message log file I saw that :

Oct 25 15:10:04 catanzaro sshd(pam_unix)[28352]: session opened for user
root by (uid=0)
Oct 25 19:12:23 catanzaro rsyncd[28269]: sent 69 bytes  received
11680096 bytes  total size 2019826667
Oct 25 15:12:52 catanzaro rsyncd[28438]: connect from
salerno.lan.lexum.pri (192.168.4.27)
Oct 25 15:12:52 catanzaro rsyncd[28438]: rsync to canlii-jur-pe from
rsync_salerno at salerno.lan.lexum.pri (192.168.4.27)
Oct 25 19:13:03 catanzaro rsyncd[28438]: sent 69 bytes  received 50125
bytes  total size 84377660
Oct 25 15:13:04 catanzaro rsyncd[28443]: connect from
salerno.lan.lexum.pri (192.168.4.27)
Oct 25 15:13:04 catanzaro rsyncd[28443]: rsync to canlii-jur-qc from
rsync_salerno at salerno.lan.lexum.pri (192.168.4.27)
Oct 25 19:14:02 catanzaro rsyncd[28443]: sent 69 bytes  received 3018179
bytes  total size 457621016
Oct 25 15:14:03 catanzaro rsyncd[28462]: connect from
salerno.lan.lexum.pri (192.168.4.27)
Oct 25 15:14:03 catanzaro rsyncd[28462]: rsync to canlii-jur-sk from
rsync_salerno at salerno.lan.lexum.pri (192.168.4.27)


As you can see there is some info from the future !

all servers are using LAN ntp server and are in sync (we are using
kerberos so time is important to us :) )

Does some of you already saw that kind of strange thinks ?