Hello everyone, I have an image larger that 2GB and I want to push to nexus registry behind traefik.
unfortunately, I have an error after trying to push that like this:
WARN [qtp971664511-699] admin org.sonatype.nexus.repository.docker.internal.V2Handlers - Error: PATCH /v2/flutter/blobs/uploads/6e08d6c2-d8b4-4071-971b-1e9ea80b3073
org.sonatype.nexus.blobstore.api.BlobStoreException: BlobId: path$nexus-repository-docker/6e08d6c2-d8b4-4071-971b-1e9ea80b3073/449115ab-a281-4e0c-b534-c3936c938d0e, org.eclipse.jetty.io.EofException: Early EOF, Cause: Early EOF
after a long time, surfing on google and read all of the blogs about this problem in traefik blogs, I didn't get a good result.
could you help me please???!!
and I have adjust this items :
- --entryPoints.web.transport.respondingTimeouts.readTimeout=3600
- --entryPoints.name.transport.respondingTimeouts.writeTimeout=3600
- --entryPoints.web.transport.respondingTimeouts.idleTimeout=3600
Are those commands on Traefik itself (i.e. static config) or the Nexus image?
Because, sadly, you cannot set up timeouts in Traefik on a per-service basis; it NEEDS to be done in the static configuration and as such will affect ALL services.
I've had the exact same use-case as you - Sonatype Nexus behind Traefik, and I was able to make it work by setting the timeouts for my HTTPS entrypoint in the static configuration file. Of course those only start to work once I restarted Traefik to reload the configuration.
Ultimately because of this "one-timeout-for-all" limitation I've decided not to use Traefik in this instance; it's not really safe in case something decides to DDoS your services...
There's some hope - it seems some people on GitHub are planning to work on this and (hopefully) fix this limitation.