Twingate fails to connect under high loads (Jenkins)

We run Twingate on our Jenkins agents, in order to connect to our Aurora clusters (PSQL), but every once in a while, it just fails to connect (Timeout).
What’s weird is that it may come back after some time (10 minutes, an hour, 3 hours) most of the times we just kill the Agent, and the next agent will probably work.

We’ve noticed that on high load the Twingate fails.

We’ve got a ticket for it, but Twingate support doesn’t help.

Any help would be great!

We found that the connection has issues from inside the docker (That runs in Jenkins)
But the agent itself is able to connect to the resource via Twingate.

Used --network host in the Docker.
Fixed the issue