On 1/03/2014 2:17 a.m., Walter H. wrote: > Hello, > > the router from ISP has IPv6 disabled; my public IP is only IPv4; > > I get regularily a message like this: > > ERROR > The requested URL could not be retrieved > ------------------------------------------------------------------------ > While trying to retrieve the URL:http://www.google.com/firefox? > <http://www.google.de/firefox?> > The following error was encountered: > > * *Connection to 2a00:1450:4001:80b::1017 Failed * > > The system returned: > /(101) Network is unreachable/ > The remote host or network may be down. Please try the request again. > Your cache administrator is my-email. > ------------------------------------------------------------------------ > Generated Fri, 28 Feb 2014 13:03:40 GMT by myproxy-host (squid/3.3.11) > > > my squid is running on CentOS 6.5 64-bit; > > I already have the following in squid.conf > > tcp_outgoing_address myproxy-host's IPv4-address all > and > dns_v4_first on The cause of this is not IPv6. Which appears to be working properly (the network is indeed unreachable). Your settings have instructed Squid to try all the IPv4 first. Squid will have dones so and found all of them failing before it even attempted the IPv6 address. Squid error page just reports the last connection failure to be seen, which happens to be on that IP. What are your settings for the following directives? balance_on_multiple_ip - if it is present, remove it. forward_max_tries - should be set to somewhere around 25 nowdays. - for 3.3 if it is missing try adding it. And what are your idns cachemgr report listings for www.google.de. PS. You could look into enabling a 6to4 tunnel connection on your proxy box. That will give your clients the use of IPv6 in these situations where IPv4 is not working. Amos
