I use Squid 3.1.19. I have four squid boxes. All of them got direct Internet Access. To share the diskcache of each Proxy, I configure them as siblings to each other.
On Squid A: cache_peer B sibling 8080 3130 proxy-only cache_peer C sibling 8080 3130 proxy-only cache_peer D sibling 8080 3130 proxy-only and of course I configure a cache_peer_access rule to prevent a request loop. For cache access I use icp. After I configure the cache_peer I got the error messages with timeout_direct to random destinations. -----Ursprüngliche Nachricht----- Von: Amos Jeffries [mailto:squ...@treenet.co.nz] Gesendet: Montag, 13. Januar 2014 22:59 An: squid-users@squid-cache.org Betreff: Re: [squid-users] TIMEOUT_DIRECT after enabling siblings On 2014-01-14 02:34, Grooz, Marc (regio iT) wrote: > Hi, > > is see some of squid request with TIMEOUT_DIRECT/IP_Address in squid > log after enabling siblings. > > Any idea? > > Kind regards > > Marc Any other details which might narrow this down? squid version(s), what you mean by "enabled siblings", whether the squid box in question has TCP issues contacting the mentioned IP, any messages appearing in cache.log, things like that could help. Amos
