Agent keeps retrying large messages when it is failing

[gnunn1]

Describe the bug

The Agent is spamming the below log message about a message being too large and stops processing other events such as updating applications. I'm not sure what is triggering this but I suspect the resource proxy since it occurs whenever I click on a resource to bring up the sidebar with the live and desired manifests. Note that this seems to happen even with a small resource.

I started my home lab this morning and after going to view a resource in the UI the agent logs are constantly getting this log message:
time="2026-02-10T16:41:11Z" level=error msg="Error while receiving: rpc error: code = ResourceExhausted desc = grpc: received message larger than max (28602193 vs. 4194304)" clientAddr="192.168.1.5:443" direction=Recv module=StreamEvent

On the Principal side I see the following messages:

time="2026-02-10T22:34:17Z" level=error msg="Timeout communicating to the agent: context deadline exceeded SUBSCRIBE {app|managed-resources|argocd-agent-hub-prod-local_cert-manager|1.8.3}" agentName=argocd-agent-hub-prod-local connectionUUID=b5ce3f8a-9bbb-4d3e-91f4-2778a35297ce function=sendSynchronousRedisMessageToAgent module=server sendEventUUID=1487b066-1ba8-4dc4-826c-80954f14ce7f
time="2026-02-10T22:34:17Z" level=error msg="unexpected nil in get response" connUUID=b5ce3f8a-9bbb-4d3e-91f4-2778a35297ce function=redisFxn module=redisProxy
time="2026-02-10T22:34:17Z" level=error msg="exit due to unable to handle agent subscribe" connUUID=b5ce3f8a-9bbb-4d3e-91f4-2778a35297ce error="unexpected nil in get response" function=redisFxn module=redisProxy
time="2026-02-10T22:34:17Z" level=error msg="Timeout communicating to the agent: context deadline exceeded SUBSCRIBE {app|managed-resources|argocd-agent-hub-prod-local_cert-manager|1.8.3}" agentName=argocd-agent-hub-prod-local connectionUUID=0bf3b4cc-f537-42b9-87c5-b57f30843a0a function=sendSynchronousRedisMessageToAgent module=server sendEventUUID=ec7ee005-9612-4013-9c11-072cd4607098
time="2026-02-10T22:34:17Z" level=error msg="unexpected nil in get response" connUUID=0bf3b4cc-f537-42b9-87c5-b57f30843a0a function=redisFxn module=redisProxy
time="2026-02-10T22:34:17Z" level=error msg="exit due to unable to handle agent subscribe" connUUID=0bf3b4cc-f537-42b9-87c5-b57f30843a0a error="unexpected nil in get response" function=redisFxn module=redisProxy

Steps to reproduce the behaviour

1. Click on a resource to view the popout with the manifests for the resource
2. (...)
3. (...)

Expected behavior

Agent should fail cleanly on a large message rather then constantly retrying.

Additional context

I'm using OpenShift GitOps 1.19 and 0.5.4 of the supported Red Hat Agent chart.