Hi, I try to set up Squid servers as web accelerator. Each Squid server should check if their siblings have the object before they get the object from the real web server. All of the object has unique URL, so it is safe to cache them forever in Squid servers. The purpose is to reduce the load on the real web server. Here is my configuration: I N T E R N E T | | +----+----+ +----+----+ | Squid 1 | | Squid 2 | +----+----+ +----+----+ 192.168.1.41| |192.168.1.40 +--------+--------+ |192.168.1.14 +-----+------+ | Web server | +------------+ My squid.conf looks like this: ##----------------------- ## squid.conf on Squid 1 ##----------------------- acl all src 0.0.0.0/0.0.0.0 acl local_dst dst 192.168.1.0/255.255.255.0 acl local_src src 192.168.1.0/255.255.255.0 http_access allow local_dst http_access allow local_src http_access deny all visible_hostname squid1.testing.com ## my host name cache_dir diskd /var/spool/squid 50000 256 256 cache_store_log /usr/local/squid/var/logs/store.log http_port 80 httpd_accel_host 192.168.1.14 httpd_accel_port 8080 cache_access_log /usr/local/squid/var/logs/access.log cache_mem 1 GB cache_effective_user squid cache_effective_group squid mime_table /usr/local/squid/etc/mime.conf client_persistent_connections off cache_log /usr/local/squid/var/logs/cache.log icp_access allow local icp_access deny all cache_peer 192.168.1.40 sibling 80 3130 refresh_pattern . 10080 100% 10080 ignore-reload override-lastmod ##---------- I checked the log files, but it doesn't seem that the Squid servers talk to their sibling. Is this the right way to do it? I tried to check the web, but I didn't find any configuration for web accelerator with siblings setting. Should I use 2 layer of Squid servers? Please advise. Thank you very much. --bk