Hi, I'm running many Splash 3.5 instances in a docker container along with HAPROXY, for load balancing, however no matter how i set the maximum number of connections or other parameters, the Splash instances all die under 24 hours since being started.
I'm attaching my config files as well as screenshot of the instances failing. what could be causing this inherent instability?
Have tried different versions of HAPROXY too but it makes no difference
the server has 34 cores and 120 GB ram so it is not an issue of resource starvation for sure
if i forcefully bring the docker down and back up again everything runs well for a time again then fails again over and over .....
what could be going wrong here?
Hi, I'm running many Splash 3.5 instances in a docker container along with HAPROXY, for load balancing, however no matter how i set the maximum number of connections or other parameters, the Splash instances all die under 24 hours since being started. I'm attaching my config files as well as screenshot of the instances failing. what could be causing this inherent instability? Have tried different versions of HAPROXY too but it makes no difference the server has 34 cores and 120 GB ram so it is not an issue of resource starvation for sure if i forcefully bring the docker down and back up again everything runs well for a time again then fails again over and over ..... what could be going wrong here?
docker-compose.yml.txt haproxy.cfg.txt