On 16/07/2015 1:51 a.m., Stakres wrote: > Hi Fred, > tests from my side: > DISKD with TCP_HIT objects: 564KB/s with wget, the same url you have tested. > AUFS with TCP_HITS objects: 47.8M/s, same wget, same squid, same url, same > all. > > Wget with AUFS: > Length: 10095849 (9.6M) [application/x-msdos-program] > Saving to: `youtube_downloader_hd_setup-2.9.9.23.exe' > 100%[======================================>] 10,095,849 47.9M/s in 0.2s > 2015-07-15 15:48:29 (47.9 MB/s) - `youtube_downloader_hd_setup-2.9.9.23.exe' > saved > > All, > We have switched some ISPs from DISKD to AUFS this morning, the "queue > congestion" appears at the begining then disappears from the cache.log. For > how long, nobody knows... Doh. Sorry its been so long I had forgotten one detail ... Squid AUFS starts with a relatively short I/O queue length (8 entries) and auto-grows it as needed by doubling the limit and displaying the congestion message. On any busy proxy this can be seen for a while after startup, then possibly during the first traffic peak after that. It is not a problem unless it keeps growing indefinitely or appears suddenly without a matching traffic spike (possible sign of disk failure). I'm applying a patch to raise the baseline queue length from 8 to 8K, and reporting in each message what the limit was changed to. That should get rid of a dozen entries in your logs, and help show whats going on in the remainder. Amos _______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users
