On Mon, 29 Mar 2010 13:58:58 -0300, Marcus Kool <marcus.kool@xxxxxxxxxxxxxxx> wrote: > I use squid > Squid Cache: Version 3.0.STABLE20 > configure options: '--prefix=/local/squid' '--with-default-user=squid' > '--with-filedescriptors=2400' '--enable-icap-client' > '--enable-storeio=aufs,ufs,null' > '--with-pthreads' '--enable-async-io=8' '--enable-removal-policies=lru' > '--enable-default-err-language=English' '--enable-err-languages=Dutch > English Portuguese' > '--enable-ssl' '--enable-cachemgr-hostname=localhost' > '--enable-cache-digests' > '--enable-follow-x-forwarded-for' '--enable-forw-via-db' > '--enable-xmalloc-statistics' > '--disable-hostname-checks' '--enable-epoll' '--enable-useragent-log' > '--enable-referer-log' > '--enable-underscores' 'CC=gcc' 'CFLAGS=-O2 -m32' 'CXXFLAGS=-O2 -m32' > <snip> > > Note that FD 27 and FD 28 have the same NODE. > This pipe is used for what ??? Something detailed in your cache.log when running at debug level ALL,1 (notices). > > The EAGAIN return code to read is strange. It suggest that the > read() could return data soon, but Squid is looping now for over 4 hours. > > Hopefully one of the developers knows what the pipe is used for and > can do a guess what is causing the EAGAIN return code. One of the helpers. Can't tell from the trace. Please also see if this is reproduced with the 3.1 release. Amos