On 02/24/2016 12:44 PM, Heiler Bemerguy wrote: > I don't think I had a bottleneck on I/O itself, In general, I/O bottlenecks with rock or ufs cache_dirs do not result in sustained 100% CPU utilization. I do not know enough about aufs to say whether that cache_dir type can have those symptoms during overload. > maybe the hash/search of cache indexes was too much for a single thread? See my earlier response regarding Squid bugs when handling huge files. Nothing you have posted so far disproves that working theory. Cheers, Alex. _______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users
