Hi all, we deploy traefik as a docker container with file configuration.
Before we had a k8s deployment with 1 pod and 2 containers inside it would communicate via traefik (side car in the pod). Now the 2 containers are in different pod and also communicating via traefik but there is one in each pod as side car
Before
Pod1(Svc1 - Traefik - Svc2)
After
Pod1(Svc1 - Traefik1) - Pod2(Traefik1 - Svc2)
Since then we started to see appearing error on Svc1 that it receives empty data from rest request (with tls). In Traefik1 access log we can see those request like that
{"ClientAddr":"127.0.0.1:43004","DownstreamContentSize":0,"DownstreamStatus":0,"Duration":114644727,"OriginContentSize":668,"OriginDuration":114431073,"OriginStatus":200,"Overhead":213654,"RequestAddr":"localhost:8046","RequestContentSize":0,"RequestMethod":"GET","RequestPath":"/xxx","RequestProtocol":"HTTP/1.1","RetryAttempts":0,"RouterName":"my-router-1-internal@file","ServiceName":"my-router-1@file","ServiceURL":{"Scheme":"https","Opaque":"","User":null,"Host":"XX.XX.XX.XX:8082","Path":"","RawPath":"","OmitHost":false,"ForceQuery":false,"RawQuery":"","Fragment":"","RawFragment":""},"StartUTC":"2023-12-15T09:31:23.870353447Z","level":"info","msg":"","origin_Content-Length":"668","origin_Content-Type":"application/json;charset=UTF-8","request_Content-Length":"0","request_Uber-Trace-Id":"e842eccf4a207bb7:6a6190384cd3897e:25c1e11eb9653fe3:0","request_User-Agent":"XX","request_X-Forwarded-Host":"localhost:8046","time":"2023-12-15T09:31:23Z"}
OriginStatus is 200 but DownstreamStatus is 0. OriginContentSize is also correct but DownstreamContentSize is 0.
This problem occurs on very rare occasion, traefik logs (2.10.1) in trace mode doesn't give more informations.