00:28:58 GMT <zpojqwfejwfhiunz> smh: that's good to hear you figured it out!
00:33:13 GMT <zpojqwfejwfhiunz> redigo has TestOnBorrow() but it's commented out for some reason so there's never a way to have it PING to make sure the connection is still present with that client
00:33:44 GMT <zpojqwfejwfhiunz> and they weren't returned to the pool because they weren't close()'d
08:43:48 GMT <smh> zpojqwfejwfhiunz: yer we implement TestOnBorrow but that doesn't help as you said the connection never gets returned to the pool
08:44:21 GMT <smh> zpojqwfejwfhiunz: I'm putting together a PR which adds KeepAlive to the default dialer which will solve it
08:44:22 GMT <zpojqwfejwfhiunz> you'd have to add something that tries a Read() periodically and sees if it's EOF
08:44:54 GMT <smh> you never get EOF for a half-closed TCP connection
08:44:55 GMT <zpojqwfejwfhiunz> what will you do for now? vendor it your fork?
08:45:03 GMT <smh> you need to actively do a write to trigger it
08:46:03 GMT <smh> yer we use glide so we can point it at a forked repo in the mean time
08:46:20 GMT <smh> garry is usually pretty good with PR's
08:46:43 GMT <smh> so should be able to get it in without too much delay
09:25:03 GMT <cloudbud> why the
09:25:04 GMT <cloudbud> hi
09:25:12 GMT <cloudbud> what is the use of timeout value
09:33:26 GMT <badboy_> cloudbud: the setting for the redis server itself?
09:41:02 GMT <cloudbud> yes
09:41:06 GMT <cloudbud> @badboy
09:42:14 GMT <badboy_> what it says
09:42:14 GMT <badboy_> # Close the connection after a client is idle for N seconds (0 to disable)
12:18:34 GMT <RedisUser> Hi there. My redis-server seems to always be at 99.7% CPU, however when I do "redis-cli" and "monitor" there isn't all that much going on. Can anyone point me in the direction to track down what might be causing the load?