I was wondering if the container image upgrade of the loadbalancer (integrated HA Proxy) would result in failed connections for the clients.
To test this I used the “ab” tool from apache to send thousands of requests to my rancher hosts (Load balancer has a running instance on every host).
When upgrading, I selected the check box to start containers before stopping the old ones, thinking that this would avoid failed requests during the upgrade process.
Unfortunately the result was not as expected, On a total of 60’000 requests that were made to the hosts, around 10’000 failed.
I would like to know if there is a better way to upgrade the load balancer (especially when upgrading the container image after rancher server upgrades).
Thanks in advance!
PS: Here is the output of the benchmarking:
This is ApacheBench, Version 2.3 <$Revision: 1807734 $> Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/ Licensed to The Apache Software Foundation, http://www.apache.org/ Benchmarking mainloadbalancer.xxx.xx (be patient) Completed 10000 requests Completed 20000 requests Completed 30000 requests Completed 40000 requests Completed 50000 requests Server Software: Server Hostname: mainloadbalancer.xxx.xx Server Port: 80 Document Path: /test Document Length: 339 bytes Concurrency Level: 1 Time taken for tests: 137.738 seconds Complete requests: 59954 Failed requests: 9780 (Connect: 0, Receive: 3260, Length: 3260, Exceptions: 3260) Non-2xx responses: 56694 Total transferred: 25285524 bytes HTML transferred: 19219266 bytes Requests per second: 435.28 [#/sec] (mean)