Hello,
I have Slackware 8.0 Linux machine (Celeron 400, 64M RAM, 128M swap)
with async 115.2k PPP link and Intel EtherExpress pro/100 ethernet with
squid-2.4.STABLE2 running
The problemm is: wheh downloading with squid browser sometimes "freeze"
on files bigger than 30-50Kb (without squid all ok). I put debug_options
ALL,2 in squid conf and found this:
2001/10/17 02:15:50| The request GET
http://mmedia.ozon.ru/graphics/img_ass/468/ozon_bibl2.gif is ALLOWED,
because it matched 'earlnet'
2001/10/17 02:15:50| clientReadRequest: FD 15: no data to process ((11)
Resource temporarily unavailable)
2001/10/17 02:15:50| fwdServerClosed: FD 16
http://u022.22.spylog.com/cnt?p=1&rn=0.05679123829377347&r=http%3A//www.
dropzone.ru/&pg=http%3A//www.dropzone.ru/guestbook.shtml
2001/10/17 02:15:55| commHandleWrite: FD 18: write failure: (32) Broken
pipe.
I understand that "Resource temporarily unavailable" and "Broken pipe"
means something is definitely wrong but not shure this is reason of my
problemm because "Resource temporarily unavailable" error after almost
every GET request even when all seem ok. "Broken pipe" happens less
frequently and also has not direct connection to freeze problemm
This is heppens when only one client connected to squid and downoad one
site at given time. This is my config
squid-2.4.stable2 and squid-2.4-200110152300 (same result)
Linux-2.4.10 and Linux-2.4.12-ac3 (same result)
After playing with squid.conf tried default (only add ACL)
No firewall, no ECN
Tried ulimit -HSn 16384 before starting squid
When I start netscape locally (on machine with squid) - the same
problemm, when I browse local site (squid, netscape and apache on same
mchine) I see same errors in cache.log but have not "freeze" problemm
(maybe just not see it)
This is some lines from cache.log:
2001/10/17 02:00:21| Starting Squid Cache version 2.4.STABLE2 for
i686-pc-linux-gnu...
2001/10/17 02:00:21| Process ID 107
2001/10/17 02:00:21| With 1024 file descriptors available
2001/10/17 02:00:21| DNS Socket created on FD 4
2001/10/17 02:00:21| Adding nameserver 212.82.205.193 from
/etc/resolv.conf
2001/10/17 02:00:21| Unlinkd pipe opened on FD 9
2001/10/17 02:00:21| Swap maxSize 102400 KB, estimated 7876 objects
2001/10/17 02:00:21| Target number of buckets: 393
2001/10/17 02:00:21| Using 8192 Store buckets
2001/10/17 02:00:21| Max Mem size: 8192 KB
2001/10/17 02:00:21| Max Swap size: 102400 KB
2001/10/17 02:00:21| Rebuilding storage in /usr/local/squid/cache
(DIRTY)
2001/10/17 02:00:21| Using Least Load store dir selection
2001/10/17 02:00:21| Set Current Directory to /usr/local/squid/cache
2001/10/17 02:00:21| Loaded Icons.
2001/10/17 02:00:21| Accepting HTTP connections at 0.0.0.0, port 3128,
FD 10.
2001/10/17 02:00:21| Accepting ICP messages at 0.0.0.0, port 3130, FD
11.
2001/10/17 02:00:21| WCCP Disabled.
2001/10/17 02:00:21| Ready to serve requests.
Could somebody please help me with resolving this problemm and
understanding error messagges?
Thanks.
Sincerely,
Elman Efendiyev
elman@earl.kharkov.ua
Received on Tue Oct 16 2001 - 18:09:00 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:02:47 MST