Can custom modules and edgeHub run on the host network?
See original GitHub issueHi there,
I’m working on a project where we have one module that communicates with several other components at the edge over a LAN which the host machine is on. We have found that the simplest way for our custom module to communicate over LAN is to configure its Docker container to use the host network. In doing so, we ran into some issues, some of which resulted from leaving edgeHub on the azure-iot-edge network. We now have edgeHub configured to use the host network as well as our custom modules, but we are still experiencing reliability issues (albeit less frequent ones). What I’m wondering is if this network configuration is even an intended use case of iotedge at this point in time, or is this undefined behavior?
Expected Behavior
edgeHub and edgeAgent are able to seamlessly work with edgeHub and custom modules on the host network instead of the azure-iot-edge bridge network.
Current Behavior
The expected behavior works but inconsistently. Sometimes edgeHub fails regularly with errors that seem to center around connectivity to IotHub. The logs attached to this post seem to indicate the failure comes from edgeHub attempting and failing to perform a twin update.
Steps to Reproduce
At this point in time, after putting both edgeHub and our custom modules on the host network there have been no reliable, reproducible failures.
Device (Host) Operating System
Ubuntu 18.04
Architecture
amd64
Container Operating System
Linux containers
Runtime Versions
1.0.2
iotedged
iotedge 1.0.2 (f6fcdb2e24ac2af977e748e9f2b8a37245b92738)
Docker
Client: Version: 18.06.0-dev API version: 1.30 (downgraded from 1.37) Go version: go1.10.2 Git commit: daf021fe Built: Mon Jun 25 21:07:53 2018 OS/Arch: linux/amd64 Experimental: false Orchestrator: swarm
Server: Engine: Version: 17.06.2-ce API version: 1.30 (minimum version 1.12) Go version: go1.8.3 Git commit: a04f55b Built: Thu Sep 21 20:36:57 2017 OS/Arch: linux/amd64 Experimental: false
Logs
Note that these logs are only a snippet of about 24 hours of logs, but the same exact error occurred again and again, so only two iterations of the failure are included for brevity. https://gist.github.com/jackt-moran/9fff63acf47e26a7ae6cb04f7ee3e4fd
Issue Analytics
- State:
- Created 5 years ago
- Comments:5 (3 by maintainers)
Top GitHub Comments
Thanks! Naturally, since posting this, I haven’t run into any issues either. If I do I will provide thorough documentation of the failure.
I’m going to close for now. I believe the underlying issue is #408