Hi Henrik,
1) I was ./configure'ing with
./configure --enable-gnuregex --enable-snmp --enable-time-hack
--enable-delay-pools --enable-dlmalloc --enable-underscores
* I tried the same on a RH7.0 and instead of habing sig violation,
squid would not respond to -k reconfigure or -k shutdown !!
* I got rid of the --enable-dlmalloc and --enable-gnuregex
and it worked well on both platform ....
* before using the delay pools, I WAS already using dlmalloc and
gnuregex
on AIX without problem ...
since after fixing the problem I didn't roll back to try adding dlmalloc
/ regex
again ... I am not 200% sure that the problem came from the simultanious
use of both...
I'll try to find a way to recreate the problem and take a trace and send
it to u guys ...
one day *S*
2) I was using a simple delay_pool with a regexp on it like the
abracadabra one from the UG.
any way, one more time, thank you for your support!
here are my new questions:
1) I read on the FAQ/UG that there was some restriction on how to use
the delay pools
based on IPv4 B/C class ...
on my network, I have dialup users, lan users and wan users( connected
through 64 / 128 lines
with different Commited Rate ).I want to give faster access to those
with bigger lines / Higher
commited rate ... I am trying to implement this wit ACL's ...
acl cust1_IP src 1.2.3.4 a.b.c.d/255.255.255.248
delay_class 1 2
delay_parameters 1 16000/8000 16000/8000 # limit to MAX 16KB/sec =
128kbps ... slow down their download after 8K
delay_access 1 allow cust1_IP
delay_access 1 deny all
am i runnning into problem using this acl with a .248 mask ? ( I have
many customer with this mask
and all should be delay'ed ) ?
is it ok to put an acl with a simgle IP + a range ?
2) I am running mrtg to monitor my server ...
is there a way to monitor how the pools are used ??
for example:
- I would like to see how full the individual buckets for WAN
customers
are so that I can sell more Commited Rate to those that
would benefit from it showing them the graphs ?
- I want to have total/ average/ means for my DialUp's to plan for
bandwith upgrade
3) I am NOT using SquidGuard ...
I was wondering how to SLOW down the sites from the blacklist
BLOCKING them ...
in a ( efficient ????) way ...
I looked on the SquidGuard intro / mailinglist archive and have not
found anything related to delay pools
so maybe there is a easy way to do that through ACL's ?
4) would someone have sample of conf file using delays based on the time
of the day ?
5) what happens when several ACLS can apply to slow the traffic down:
ex: - a users is limited to 36.6 kpbs by the 1st delay pool
- he is downloading something from a warez site
would the 2 buckets fill up ?
Henrik Nordstrom wrote:
>
> A stack trace of where it crashes would help a lot in identifying the
> problem.
>
> Also, what delay pool settings are you using?
>
> --
> Henrik Nordstrom
> Squid hacker
>
> Fabrice Laborie wrote:
> >
> > Hi Guys,
> >
> > has anyone been using delay pools on AIX ?
> >
> > I get a Seg Violation when I start my server :(
> >
> > 2000/10/28 17:54:43| Ready to serve requests.
> > FATAL: Received Segment Violation...dying.
> > 2000/10/28 17:54:43| Not currently OK to rewrite swap log.
> >
> > note:
> >
> > * I've been running squid for 2 months now ...
> > * I have just ./configured with --enable-delay-pools ( after a make
> > distclean)
> > * if all delay configurations are commented out (#) the server runs
> > normally
> > * it crashes as soon as I uncomment my delays settings in the squid.conf
> >
> > ps: I run 2.3STable4 on AIX 4.3.2 compiled with gcc 2.95.2 from bull
> >
> > --
> > To unsubscribe, see http://www.squid-cache.org/mailing-lists.html
-- To unsubscribe, see http://www.squid-cache.org/mailing-lists.htmlReceived on Mon Oct 30 2000 - 23:35:09 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:56:03 MST