>We have two important regional caches running on PC/Linux (I believe it's
>2.0) and the Korean national cache uses it too.
yes, I looked at the korean one. I see they are a linux 2.0.x kernel along
with the 1.0.17 squid.. I would guess that they are probably getting "funny"
ip numbers in their cache and access logs or they modified their code.
BTW, we have found that most of the errors that we are having are not due to
linux, but due to the clients. In particular there *seems* to be a yet as
undiscovered problem with the Distinct TCP/IP stack which *seems* to be
causing a high number of "connection resets" (a high percentage of our users
are using the Distinct stack).
Has anyone out there done a count of connection resets in cache log/lines in
access log.. We are running at about 10%.. Anybody else have any figures.
jon zaid
Received on Fri Nov 29 1996 - 12:04:52 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:33:41 MST