In article <0EPIURP8L0036Z@mailman.otis.com> you write:
> redirector #1: 159079 (159079 rewrites)
> redirector #2: 7794 (7794 rewrites)
> redirector #3: 1760 (1760 rewrites)
> redirector #4: 506 (506 rewrites)
> redirector #5: 168 (168 rewrites)
>
>Which looks OK to me. I'm not sure why there's a discrepancy of one
No...
>between the requests and the replies (presumably this is because the
>cachemgr "cache_object" query is still in progress and thus not fully
>replied to - anyone comment / is this a minor bug that could be fixed ?)
>
>Since redirector 5 is getting a bit of a work out, should I increase the
>numbers a bit, so that there are always a few "spare" redirectors for
>when there are more than 5 concurrent requests needing redirectors to
>respond. Is there a FAQ / rule of thumb on how to "size" the number of
>redirect_children in squid.conf ? Anyone want to share there
>experiences with me ?
Just make sure there are always a few redirectors/dnsservers with 0 rewrites
left. I'm not sure if it has something todo with the problem you're seeing.
Arjan
Received on Tue Mar 10 1998 - 13:03:58 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:39:14 MST