I have a 32bit pentium 4, running sxce build 130. 2 GB of memory,
using a ZFS mirror on local PATA disks.
Under TCP load (about 70mbps sustained), I see the following output
when running 'netstat -i 1':
l...@bfd-2: netstat -i 1
input iprb0 output input (Total) output
packets errs packets errs colls packets errs packets errs colls
54300013 0 26903555 1 0 54300275 0 26903817 1 0
8537 0 4270 18446744073709551615 0 8537 0 4270
18446744073709551615 0
5780 0 2892 0 0 5780 0 2892 0 0
8778 0 4390 0 0 8778 0 4390 0 0
That massive number of errors shows up periodically. Obviously,
there's no way to collect that many real errors in one second.
I get this error several times a minute, and the error count is always
the same. I note that 18446744073709551615 happens
to be (2*1024^6)-1.
Now that seems really suspicious to me, like some 64bit<>32bit
math error is happening.
Has anyone else seen this with a iprb interface?
-Kurt
--
This message posted from opensolaris.org
_______________________________________________
networking-discuss mailing list
[email protected]
