Hi Guys,
We have been using haproxy in our company for months and have loved our
experience. However we recently ran into a strange issue even though no changes
happened on the haproxy side.
We are currently using haproxy 1.4.24 on Debian 7. Our backend consists of 6
Windows 2008 R2 servers running IIS 7.5. Strictly web servers for stateless
applications and services round robin balanced.
I have been debugging with fiddler and usually requests work fine however once
every few requests a page will hang and it appears that haproxy makes a connect
to a backend server but the data stream is lost before all the data is
transferred.
Within Fiddler I receive a content length mismatch errors saying we should have
downloaded 800 KB but only received 100KB. Our client and server connections
are set to 10 min and the connect timeout is 5 seconds (increasing that had no
effect).
I am able to recreate the issue but it is random and is not related to server
load or network limitation as we don't have any other issues. I recreate by
refreshing a page over and over until the issue arises.
If I point my browser to a backend server around haproxy the issue does not
happen. Very perplexing...
Below find config...
Global config
global
log 127.0.0.1 local0
log 127.0.0.1 local1 notice
maxconn 100000
stats socket /var/run/haproxy.sock mode 0600 level admin user haproxy group
haproxy daemon
Default config
defaults
log global
mode http
option httplog
option dontlognull
option redispatch
option forwardfor
option httpclose
option abortonclose
retries 1
timeout connect 5000
timeout client 50000
timeout server 50000
listen stats
#disabled
bind *:8888
stats enable
stats uri /haproxy?stats
stats realm Strictly\ Private
stats auth xxxxx:xxxxxx
frontend open *:80
timeout client 600000
default_backend web
backend web
timeout server 600000
balance roundrobin
server edrwebproawvc1 xxx.xxx.xxx.xxx:80 check
server edrwebproawvd1 xxx.xxx.xxx.xxx:80 check
server edrwebproawvc2 xxx.xxx.xxx.xxx:80 check
server edrwebproawvd2 xxx.xxx.xxx.xxx:80 check
server edrwebproawvc3 xxx.xxx.xxx.xxx:80 check
server edrwebproawvd3 xxx.xxx.xxx.xxx:80 check
I'm curious if you have seen similar issues before and what your thoughts are
that could be causing the issue.
Also what is the best way to log these kinds of issues. In my http log for
haproxy I see nothing when these issues happen because I think the haproxy
connections are alive it is just the data stream that is lost.
Any help you can point me to I would greatly appreciate it!
John Dzialo | Linux System Administrator
Direct 203.783.8163 | Main 800.352.0050
Environmental Data Resources, Inc.
440 Wheelers Farms Road, Milford, CT 06461
www.edrnet.com<http://www.edrnet.com/> |
commonground.edrnet.com<http://commonground.edrnet.com/>
[Description: Description: Description: Description: Description:
EDR_logo4color_EDR_only_80px2]
