I have AWS configured to run farms within a VPC in scalr without problems. AWS Governance has the VPC
policy set to "Allowed" so that when creating GCE farms they seem be set to VPC by default (I didn't see an
option to create Governance rules for GCE). In the file scalr-server.rb I have:
# amazon specific
:ec2 => {
# use private IP for communicating with instances
:instances_connection_policy => 'local',
},
# google specific
:gce => {
# use private IP for communicating with instances
:instances_connection_policy => 'local',
},
When launching the GCE farm, I get pass Provision Server in Cloud Platform
but it gets stuck inside Wait for OS to finish booting. I ssh into the instance
and found the following message at the bottom of syslog:
May 7 19:16:19 ccc122d6-277d-4875-816a-e4e37729052e google: No startup script found in metadata.
I've tried changing :instances_connection_policy => 'auto'
and
app[:instances_connection_policy] = 'local'
but this is not helping. I can ping the scalr server within the gce instance and vice-versa so
why the GCE instance gets stuck in the scalr OS booting stage?