Issue with docker swarm when high traffic

Open Source Projects Swarm
1

Hi support team!

Current we’re running docker swarm with 4 node :

  • 3 node master + 1 node worker.
  • All 4 node docker server running VM with mode global
  • Os : Ubuntu 16.04.5 LTS
  • Docker version : 18.09.0
  • We use gitlab to run CI/CD deploy service via docker stack with healthcheck
    Occasionally we encounter problems like this:
  • When one node in docker swarm load very high and can’t resolve and other node cpu usage very low.
  • We tried kill container and when container healthy but it continue to load high and experience endless looping.

Anyone ever encountered this problem and how to resolve it.
Please help me!

Thanks!

2

You mean: Hi community!
If you expect the “support” team to respond, you will need to buy a subscription or raise an issue in docker’s github project.

Btw. with this level of provided details, the best thing you will get here is a wild guess. You need to gather more information regarding the overall state and the problem. Make sure to check your log management system (loki? ELK? Graylog? Splunk?) and your system monitoring (Prometheus/Grafana? Icignia? .)