TCP error messages in 6.0
Patrik Roos
darkgray at gmail.com
Thu Nov 10 03:49:43 PST 2005
After 6.0 hit stable, I decided to upgrade my machine. Uname now reads:
FreeBSD pumpkin.lolikon.org 6.0-STABLE FreeBSD 6.0-STABLE #0: Sat Nov
5 08:12:50 CET 2005
darkgray at pumpkin.lolikon.org:/usr/obj/usr/src/sys/PUMPKIN i386
After the upgrade, however, I started getting spurts of error messages
posted to me in the daily security run output, looking like this:
pumpkin.lolikon.org kernel log messages:
+++ /tmp/security.rHb8JAYI Thu Nov 10 03:02:41 2005
+tcp_sack_output: Computed sack_bytes_retransmitted (2904) not the
same as cached value (5808)
+tcp_sack_output: Computed sack_bytes_retransmitted (2864) not the
same as cached value (5768)
+tcp_sack_output: Computed sack_bytes_retransmitted (2904) not the
same as cached value (5808)
+tcp_sack_output: Computed sack_bytes_retransmitted (2904) not the
same as cached value (5808)
+tcp_sack_output: Computed sack_bytes_retransmitted (2896) not the
same as cached value (5800)
These appear with a timestamp in /var/log/messages, and it seems I get
800 of these in one second, once a day at random timestamps. Never saw
them in 5.4, so I'm assuming it's because of the 6.0 kernel.
This is my ifconfig report on the NIC:
xl0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
options=9<RXCSUM,VLAN_MTU>
inet6 fe80::210:4bff:fe07:279d%xl0 prefixlen 64 scopeid 0x1
inet 192.168.0.3 netmask 0xffffff00 broadcast 192.168.0.255
ether 00:10:4b:07:27:9d
media: Ethernet autoselect (100baseTX <full-duplex>)
status: active
Any ideas on cause and solution? I've never posted on this list
before, so if you manage to help, it'll make a very good first
impression. ;)
Regards,
Patrik Roos
More information about the freebsd-stable
mailing list