Closed legolas0802 closed 1 month ago
Does it happen every 10 minutes? It is possible the proxy container closes that connection every 10 minutes ... is that something you can check? I myself I am not using the proxy container and it looks like you are the first seeing this problem.
I check and let you know, but if the container closes the connection, shouldn't all the entities disappear? general docker information and switches remain there and are fully functional
i checked and it happens every 10 minutes
Ok, confirmed: https://github.com/Tecnativa/docker-socket-proxy/blob/master/haproxy.cfg
It has: timeout client 10m timeout server 10m
Your connection which closes, is the docker event thread ... so it only gets traffic when a container stops, starts, etc
You can download the haproxy.cfg from the GitHub link (above), modify it (e.g. 10min to 60min), and mount it inside the container.
Other option could be: configure a container with "health" check, this should generate event traffic. I cannot do this from monitor_docker.
btw if you run "docker events" you can see the events push from docker on the screen.
I tried to increase the container time to 60 minutes, for both the client and the server. The problem is only postponed
I did another test, deleting the entries:
timeout client 10m timeout server 10m
the problem no longer arises.
I am not sure you do not get any "docker events", because those should keep the connection open (aka alive). But thanks for further testing, I will add your fix to the README.
sorry if the question is rather stupid, but I have the same problem. however, my docker knowledge is quite poor, can I edit the haproxy via portainer? if not, can you help me how to get these two lines out?
Thanks a lot!
scusate se la domanda è un po' stupida, ma ho lo stesso problema. tuttavia, la mia conoscenza della finestra mobile è piuttosto scarsa, posso modificare l'haproxy tramite portainer? se no, puoi aiutarmi come ottenere queste due righe?
Molte grazie!
through portainer enter the console of the proxi container in / bin / sh mode, move to the path usr / local / etc / haproxy / haproxy.cfg and with the command "vi" edit the file
guys, is deleting lines
timeout client, timeout server
a true workaround solution? I'm using docker-socket over tcp connection and got the same problem with ha logs
I will close this old issue. If still persists, please open a new issue.
I am on home assistant supervisor and I use the proxy container as per instructions, in the log I find the following error:
with the consequence that those entities disappear for all containers, to see them reappear after a few minutes. All in an infinite cycle.