On 16/07/2015 2:59 a.m., Yuri Voinov wrote: > > Amos, > > I think, auds queue must be buffered more better and smoother. On some > OS (I've tested) peak loads leads performance degradation. Periodically. > Buffering and I/O scheduling is all done by the system disk controller AFAICT. Squid is just accumulating a queue of I/O requests while it waits for the controller to signal that its ready for them. Then that lot get sent to it. Et, loop.. It may be disk driver related. Or on the other end if Squid is too busy processing other stuff when the FD signals come back from the disk controller indicating the read()/write() API calls need to happen to pick the data up. Its running on libpthreads, which I know has issues on some systems. But theres nothing we can do about that bit, and thing are improving slowly anyway. Amos _______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users