Hello, On Thu, Apr 20, 2017 at 11:03:42PM +0800, jaseywang wrote: > Our haproxy 1.5.4 MS cluster performs quite well before, and the peak (...) > Now, the weird thing is why haproxy has so many closewait connections? and > why the backlog queue soon becomes full? Usually so many closewait means > haproxy can't correctly close the connection.
The first explanation I'm seeing is here : http://www.haproxy.org/bugs/bugs-1.5.4.html Your version is totally outdated, contains at least 183 well known bugs which have since been fixed in the same branch, out of which 18 are considered as MAJOR (ie severe impact) and 88 medium (ie: severe impact but with possible workaround). Is there a reason you never found the time to keep it up to date during the last 2.5 years ? We maintain stable branches exactly so that users never face such issues without being forced to upgrade to a new branch, despite this it seems that once in a while some people get trapped, it's really sad :-( Willy
