Copy Cluster between two similar cluster

Joseph

We have planned to upgrade os from cent0S6 to cent0S7 and we will use copy cluster between two similar cluster procedure (Vertica version 9.1).
I have few questions-
A)We will keep the source cluster and its etl and reporting running, will their performance be impacted?
B)What happens to transactions that are done during copy cluster,will that data will be present in target cluser?
C)We are already doing object level backup....i think full backup is not necessary before copy cluster...as anyways the source cluster will always be up?

Joseph

@mosheg @Jim_Knicely or anyone can you please clarify.

mosheg

a) We will keep the source cluster and its etl and reporting running, will their performance be impacted?
If your system is I/O bound and overutilized then yes. If your system has enough nodes then the impact will be low.

b) What happens to transactions that are done during copy cluster, will that data will be present in target cluster?
Copycluster uses the Linux rsync tool to copy files from the source cluster to the target cluster.
All committed data at the time vbr determined what data to copy will be presented in the target cluster.
After you use copycluster, you can then use the replicate task to update tables and schemas instead of doing another full copy.
The first replication of an object replicates the entire object. Subsequent replications copy only data that has changed since the last replication.
See: https://www.vertica.com/docs/10.1.x/HTML/Content/Authoring/AdministratorsGuide/BackupRestore/ReplicatingtoanAlternateCluster.htm

c) We are already doing object level backup....i think full backup is not necessary before copy cluster...as anyways the source
cluster will always be up?
Yes, when using the --task copycluster switch the source database must be up, and the target database must be down.
Copying your database to another cluster using copycluster, is essentially a simultaneous backup and restore operation. The data is backed up from the source database cluster and restored to the destination cluster in a single operation.