Page 1 of 1

replacing an instance

Posted: Thu Feb 04, 2021 1:25 pm
by BanditBBS
So I have an existing 2 node cluster with both running Linux 6.10. I want to deploy a new host running RHEL 8 and remove one of the node and replace with the Linux 8 new build. I can find the instructions to add a host into a cluster, but I can't find option to remove an existing node from a cluster.

How do I do that? Is my plan the best way to accomplish upgrading the OS?

Re: replacing an instance

Posted: Fri Feb 05, 2021 11:16 am
by cdienger
Yes, this would be the best path to upgrade the OS.

Keep in mind that adding the new node to the existing cluster may exceed the license limit and the dashboards section of NLS will not be available. Data will still be accepted and inserted into the database however.

Once the new node is synced and the status of the cluster is green you can remove the old node following https://assets.nagios.com/downloads/nag ... luster.pdf. The dashboards section will become available again when the number nodes in the cluster no longer exceeds license limits.

I will also point out that closed indices will not be synced with the new node. If you need to sync them then they will need to be reopened. See https://github.com/elastic/elasticsearch/issues/12963 for a script to automatically open, wait for them to sync, and then close them again.

Re: replacing an instance

Posted: Tue Feb 23, 2021 12:56 pm
by BanditBBS
Trying to do this now and when adding nod to cluster I get: Could not generate node_uuid - check file permissions

Code: Select all

[root@cms-chi-nls01 tmp]# ls -l /usr/local/nagioslogserver/var
total 16
-rwxrwxr-x. 1 nagios nagios  0 Feb 22 11:32 auditlog.log
-rwxrwxr-x. 1 nagios nagios 10 Feb 22 11:32 cluster_hosts
-rw-rw-r--. 1 nagios nagios 37 Feb 22 11:35 cluster_uuid
-rw-r--r--. 1 nagios users  90 Feb 23 11:00 hosts.log
-rw-r--r--. 1 apache apache  0 Feb 23 11:54 jobs_apache.log
-rw-r--r--. 1 nagios users   0 Feb 23 11:54 jobs.log
-rw-rw-r--. 1 nagios nagios 37 Feb 22 11:34 node_uuid
-rw-r--r--. 1 nagios users   0 Feb 23 11:54 poller.log

Re: replacing an instance

Posted: Tue Feb 23, 2021 1:03 pm
by BanditBBS
Seems to be a DNS resolution issue on the host. used IP and its working instead of hostname. Now it cant contact the ldap server had to login with the local nagiosadmin. So disregard for now.

Re: replacing an instance

Posted: Tue Feb 23, 2021 5:45 pm
by scottwilkerson
BanditBBS wrote:Seems to be a DNS resolution issue on the host. used IP and its working instead of hostname. Now it cant contact the ldap server had to login with the local nagiosadmin. So disregard for now.
Sounds good

Locking thread

Feel free to open another if this is still an issue