I didn't catch this when it happened, but I noticed it later when starting
a tcpdump to track something else.
Anyway, here is a sample:
Notice the totally bogus sack entry fire is sending out (it's sending the
same one over and over again, hasn't stopped yet), but also notice how
close it is to wrapping around 4g, so there is probably a slighly bogus <
check that thinks the sack isn't outdated:
[root@fire finnag]# tcpdump -s 128 host gate.guardian.no and port 4242
06:42:28.621448 gate.guardian.no.62104 > fire.pvv.org.4242: P 1722019169:1722019171(2) ack 1720562704 win 31856 <nop,nop,timestamp 271772 4295860> (DF)
06:42:28.621448 fire.pvv.org.4242 > gate.guardian.no.62104: . ack 2 win 31856 <nop,nop,timestamp 4297071 271772,nop,nop, sack 1 {4294967163:4294967221} > (DF)
06:42:28.881448 fire.pvv.org.4242 > gate.guardian.no.62104: P 1:124(123) ack 2 win 31856 <nop,nop,timestamp 4297097 271772,nop,nop, sack 1 {4294967163:4294967221} > (DF)
06:42:28.881448 fire.pvv.org.4242 > gate.guardian.no.62104: P 124:140(16) ack 2 win 31856 <nop,nop,timestamp 4297097 271772,nop,nop, sack 1 {4294967163:4294967221} > (DF)
06:42:28.991448 gate.guardian.no.62104 > fire.pvv.org.4242: . ack 140 win 31856 <nop,nop,timestamp 271809 4297097> (DF)
In the other end that became
06:42:28.602272 lucifer-10MB.guardian.no.1066 > fire.pvv.org.stuffd: P 1722019169:1722019171(2) ack 1720562704 win 31856 <nop,nop,timestamp 271772 4295860> (DF)
06:42:28.669362 fire.pvv.org.stuffd > lucifer-10MB.guardian.no.1066: . ack 2 win 31856 <nop,nop,timestamp 4297071 271772,nop,nop, sack 1 {4294967163:4294967221} > (DF)
06:42:28.952661 fire.pvv.org.stuffd > lucifer-10MB.guardian.no.1066: P 1:124(123) ack 2 win 31856 <nop,nop,timestamp 4297097 271772,nop,nop, sack 1 {4294967163:4294967221} > (DF)
06:42:28.963552 fire.pvv.org.stuffd > lucifer-10MB.guardian.no.1066: P 124:140(16) ack 2 win 31856 <nop,nop,timestamp 4297097 271772,nop,nop, sack 1 {4294967163:4294967221} > (DF)
06:42:28.966647 lucifer-10MB.guardian.no.1066 > fire.pvv.org.stuffd: . ack 140 win 31856 <nop,nop,timestamp 271809 4297097> (DF)
- Finn Arne
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu