Cause:
An additional required network has been added after deploying the first host.
Consequence:
hosted-engine-setup fails trying to deploy an additional host
Fix:
Adding the host in a non-operational status. Printing a hint for the user that needs to configure network bindings for that host.
Description of problem:
When adding a new host to the hosted-engine setup, if there is another network in the cluster with the required flag - which I believe is default for any new networks required.
The new host will stay in and eventually time out. This is because the engine has come up within the engine as non-operational as it doesn't meet the cluster's network requirements. (ie. a required network isn't configured on the host)
[ INFO ] Still waiting for VDSM host to become operational...
[ INFO ] Still waiting for VDSM host to become operational...
[ INFO ] Still waiting for VDSM host to become operational...
[ INFO ] Still waiting for VDSM host to become operational...
[ INFO ] Still waiting for VDSM host to become operational...
Logging into the engine and assigning the correct networks will let the host come up as operational and succeed the install.
Version-Release number of selected component (if applicable):
How reproducible:
Always
Steps to Reproduce:
1. Install first host
2. Configure a new network in the Default cluster
3. Attempt to bring up a second host
Actual results:
Will timeout
Expected results:
Should let the host be successfully installed and let the user add the networks later. Possibly have it join the cluster in maintenance mode to avoid the issue?
Additional info:
Comment 1Simone Tiraboschi
2014-10-28 16:00:52 UTC
Reproduced, I'm attaching the relevant log files as a reference for further investigation.
Comment 2Simone Tiraboschi
2014-10-28 16:05:01 UTC