VCH Deployment Fails with Docker API Endpoint Check Failed Error
When you use
vic-machine create to deploy a virtual container host (VCH), deployment fails because
vic-machine cannot contact the Docker API endpoint.
Deployment fails with with the error:
Docker API endpoint check failed: API may be slow to start - try to connect to API after a few minutes: Run docker -H 192.168.218.160:2376 --tls info If command succeeds, VCH is started. If command fails, VCH failed to install - see documentation for troubleshooting.
vic-machine checks that the endpoint VM is reachable from Docker clients. If this check fails,
vic-machine create fails with an error. This error can be caused by the Docker API being slow to start or because it has failed to start.
The solution to choose depends on whether the API is slow to start or whether it failed to start.
Docker API is Slow to Start
Wait for a few minutes, then run the
docker info command to test the responsiveness of the Docker API.
docker info succeeds, it shows information about the VCH, including confirmation that the storage driver is vSphere Integrated Containers.
Storage Driver: vSphere Integrated Containers version Backend Engine
This output means that the VCH is running correctly and can now accept Docker commands.
docker info times out, it means that the Docker API did not start.
Docker API Did Not Start
If the Docker API was not responsive when you ran
docker info, download the VCH log bundle and examine the logs to determine why the deployment failed. Collecting the vSphere log bundle might also be useful for troubleshooting.
- For information about how to download VCH logs by using the VCH Admin Portal, see Access the VCH Admin Portal in vSphere Integrated Containers Engine Administration.
- For information about how to collect VCH logs manually, see Access vSphere Integrated Containers Engine Log Bundles in vSphere Integrated Containers Engine Administration.