CJE cluster-init fails when configured for internal VPC on AWS

Issue

  • When creating a new CJE cluster that uses internal VPC, the cluster-init operation fails with error relating to access to Mesos

  Error querying mesos metrics at http://mesos.xr-jenkins-controller-239852362.us-east-1.elb.amazonaws.com.elb.cloudbees.net/metrics/snapshot: <urlopen error timed out>
Mesos failed to start after 40 attempts.
There were one or more errors

??

Environment

??

Resolution

When installing CJE to an internal VPC on AWS there are several items you must provide in the cluster-init.config

vpc_id : ID of the existing VPC
vpc_subnet_id : ID of the existing subnet
additional_security_group_id : ID of an existing security group ID that will be added to all instances and elb of the cluster.
internal : must be set to yes

Errors or misconfigurations associated with these values can cause the verifications to fail and then fail the init operation.

See addtional detail on internal VPCs and refrence architecture in this section of the CJE admin guide

Have more questions? Submit a request

0 Comments

Please sign in to leave a comment.