Syslog duplicate message suppression seems to be eating more messages than it should.

[bga]

I was starting to track down a bug related to query update messages being lost and I started it by checking the syslog for any suspicious messages. Here is the relevant part of what I saw:

2009-07-20 19:45:33 DAEMON 'DHCP': DHCP timeout shift: 4 secs (try 0)
2009-07-20 19:45:37 DAEMON 'DHCP': DHCP timeout shift: 8 secs (try 0)
2009-07-20 19:45:45 DAEMON 'DHCP': DHCP timeout shift: 2 secs (try 1)
2009-07-20 19:45:46 KERN: bfs: bfs_open_dir:1550: Invalid Argument
2009-07-20 19:45:47 DAEMON 'DHCP': DHCP timeout shift: 4 secs (try 1)
2009-07-20 19:45:47 KERN: Last message repeated 2577 times.
2009-07-20 19:45:51 DAEMON 'DHCP': DHCP timeout shift: 8 secs (try 1)
2009-07-20 19:45:59 DAEMON 'DHCP': DHCP timeout shift: 2 secs (try 2)
2009-07-20 19:46:01 DAEMON 'DHCP': DHCP timeout shift: 4 secs (try 2)
2009-07-20 19:46:05 DAEMON 'DHCP': DHCP timeout shift: 8 secs (try 2)

The interesting line is the "Last message repeated 2577 times". if you look at the syslog it seems to be saying that the message above it was repeated this many times but it was obviously not it. This message was a result of deleting 2577 emails at once from Tracker so it seems the syslog not only ate the repeated messages but also at the first message that it should show (which I am pretty sure is "port full" or something like that).

[mmlr]

This looks perfectly valid to me. The repeated line in this case is:

2009-07-20 19:45:46 KERN: bfs: bfs_open_dir:1550: Invalid Argument
2009-07-20 19:45:47 KERN: Last message repeated 2577 times.

Note that the DHCP messages aren't written by the kernel, they are inserted by the net_server using the syslog() command.

[bga]

Makes sense, but this is confusing. Anyway, closing as invalid.