Question related to changing cluster RF2 to RF3


Userlevel 1
Badge +16
Hi everybody,
I have got some questions related to cluster RF operation that needs support from you all
Let me list them in sequence below:
1. How can I verify the time taken to change from RF2 to RF3 in a 5node cluster?2. While changing from RF2 to RF3, can I stop this progress and then resume the task later?3. How can I know the RF3 changing progress has been completed or not?4. Can we monitor the progress of RF2 to RF3?5. When VM are running on the cluster, does its performance get impacted anddegraded? In the case of degraded VM, how much ratio is it?
6. Will the throughput of copying data of VM in the RF2 to RF3 progress be run by an constant speedWill this speed can be modifed? and what is this speed depend on?7. Assume that a VM has 100GB data, how long will it take to complete copyingdata from RF2 to RF3?

Thanks in advance

7 replies

Userlevel 1
Badge +16
Anyone can help me on this?

Userlevel 4
Badge +20
Hi hienle

There is two parts to the upgrade. the metdata goes to RF5 to support lossing two nodes. After that is complete you can select which containers you want to go to RF3. Metadata uses dynamic ring changer to perform metadta RF Migration scans to increase the RF.

On the data side the rebuild process happens with curator is throttled. Roughly around 40 MB/s per node for hybrid and around 75 MB/s per node for all flash. For hyrbid we rebuild HDD to HDD so cold data doesn't hurt performance.




Userlevel 1
Badge +16
dlink7
Thanks a lot for your information
By the way, may I have some command tools or GUI that I can monitor the progress?
And I really do want to know the time taken to do the RF change, how to control the progress, and how to verify the speed of data rebuild process?

Really appreciated

Userlevel 4
Badge +20
The ctr going from RF2 to RF3 progress can be seen from the disk build page on

The resliency status will change on the "Data Resiliency Status" on the Prism home page.
Userlevel 4
Badge +20
The ctr going from RF2 to RF3 progress can be seen from the disk build page on cvm_IP:2010

The resliency status will change on the "Data Resiliency Status" on the Prism home page.
Userlevel 1
Badge +16
dlink7

Thank you a lot.
Can you share with me how to see the RF2 to RF3 on the disk build page?
I have accessed to that page but really did not know what happened there.
I am sorry, I quite new to Nutanix so...
Userlevel 1
Badge +16
dlink7
I have attached the screenshot as below:



After the 2 Jobs above completed, may I come into a conclusion that:
- The 100GB data of VM running inside of the RF3 cluster has been replicated into RF3 successfully?
- The RF3 changing progress has been completed successfully?
- While those 2 jobs running, may I stop them?
It is a cluster running with 8 nodes. The container that I made to be changed to RF3 has only one VM that 100GB data written

Thanks in advance

Reply