Cloud Volumes ONTAP

write-prevent > 70%, in FabricPool?

netappmagic
2,156 Views

In FabricPool, ontap 9.5p3, Writes from the capacity tier to the performance tier are disabled if performance tier capacity is
greater than 70%. If this occurs, blocks are accessed directly from the capacity tier. 

 

Can I change or increase 70% the parameter?

Are there anyway I can tell if the data is being pulled back? because I wanted to monitor the data flow and performance.

Anybody has experiences the speed to write data back, if the data got accessed?

 

Thanks

1 REPLY 1

anee
1,034 Views

https://www.netapp.com/pdf.html?item=/media/17239-tr-4598.pdf -page 15

Write-back prevention
If the local tier is at >90% capacity, cold data is read directly from the cloud tier without being written back to the local tier. By preventing cold data write-backs on heavily utilized local tiers, FabricPool preserves
the local tier for active data.


Prior to ONTAP 9.7, write-back prevention took place when the local tier was at 70% capacity.


From Blue XP :  this can be adjusted as below :

https://docs.netapp.com/us-en/bluexp-cloud-volumes-ontap/task-tiering.html#changing-the-free-space-ratio-for-data-tiering


Are there anyway I can tell if the data is being pulled back? because I wanted to monitor the data flow and performance.
=> During data access a performance archive may include counters to tell if data is read from local tier or cloud tier.

If read by random reads, cold data blocks on the cloud tier become hot and are moved to the local tier. If read by sequential reads such as those associated with index and antivirus scans, cold data blocks on the cloud tier stay cold and are not written to the local tier.

Performance support ticket can help gather these details during data access to understand above.

 

 

Anybody has experiences the speed to write data back, if the data got accessed?
https://www.netapp.com/pdf.html?item=/media/17239-tr-4598.pdf
Page 32

Cloud retrieval

On-Read:
• When cold blocks in a volume with a cloud retrieval policy set to On-Read are read, randomly or sequentially, they are made hot and written to the local tier.
• Applications that use sequential reads triggers write-backs to the local tier by setting the volume cloud retrieval policy to On-Read. This can be beneficial for applications that need local-tier performance from previously cold data that is now being read by active workloads.
Promote:
• Setting the cloud retrieval policy to Promote immediately queues tiered data to return to the local
tier—provided the tiering policy allows it. For example:




Public