Support Questions

Find answers, ask questions, and share your expertise
Announcements
Celebrating as our community reaches 100,000 members! Thank you!

Redeploying ManagerNode?

avatar
New Contributor

My ManagerNode has gotten corrupted - I can get to the CM UI, but can't SSH in (server refuseed our key) I've not changed any config, and reboot doesn't fix.

Is there anyway of re-rolling just the ManagerNode?

 

Cheers,

 

PC

 

1 ACCEPTED SOLUTION

avatar
Guru
Paul,

It is possible to repair the instance by mounting it's EBS volume on
another instance and copying the key files from another working instance in
the cluster (since they all accept the same key). You would need to do this
manually, however - It's not something Cloudera Live can do. It has no
direct access to your account. It works by SSH credentials set up by the
instances, and then deletes these keys when it's finished.

There are several blog and forum posts detailing how to do this, e.g.
http://automagical.rationalmind.net/2015/09/28/regain-access-aws-ec2-instance-lost-access/.
It is a tricky procedure. If you don't have valuable data on this cluster,
you may also consider deleting your cluster, registering for a new access
code and starting a new cluster.

View solution in original post

1 REPLY 1

avatar
Guru
Paul,

It is possible to repair the instance by mounting it's EBS volume on
another instance and copying the key files from another working instance in
the cluster (since they all accept the same key). You would need to do this
manually, however - It's not something Cloudera Live can do. It has no
direct access to your account. It works by SSH credentials set up by the
instances, and then deletes these keys when it's finished.

There are several blog and forum posts detailing how to do this, e.g.
http://automagical.rationalmind.net/2015/09/28/regain-access-aws-ec2-instance-lost-access/.
It is a tricky procedure. If you don't have valuable data on this cluster,
you may also consider deleting your cluster, registering for a new access
code and starting a new cluster.