Cost incurred by Enabling cool access on ANF Premium and Ultra service level

Question

It is mentioned in the document that

"The QoS limits on Premium and Ultra service level capacity pools enabled with cool access
will be updated in the future, with throughput on:

• Premium pools capped at 36 MiB/s per TiB (instead of 64 MiB/s per TiB) and
• Ultra pools capped at 68 MiB/s per TiB (instead of the 128 MiB/s per TiB)"

where exactly is the throughput being limited is it at the specific volume(s) which is enabled with cool-access which would impacted or is it the capacity pool (all volumes in that capacity pool)

In this case, the expected throughput on a specific volume would almost drop by 50%.
Wouldn't that cause performance issue? Since QoS is already half of what is expected at a premium service level, to achieve the usual throughput for the data in hot tier, wouldn't we to overprovision the volume in such case? Just to achieve the usual throughput on a normal premium service level.

How would it be cost efficient?

Answer 1

Hi Duggu,

The throughput limits you mentioned for Azure NetApp Files Premium and Ultra service levels with cool access apply to the entire capacity pool, not just specific volumes

This means that all volumes within a capacity pool enabled with cool access will experience the reduced throughput limits.

Regarding your concern about performance issues, it's true that the expected throughput on a specific volume would drop by approximately 50% when cool access is enabled. This reduction can indeed impact performance, especially for workloads that require high throughput. To achieve the usual throughput for data in the hot tier, you might need to overprovision the volume, which could lead to increased costs.

However, enabling cool access is designed to be cost-efficient for scenarios where a significant portion of the data is infrequently accessed (cold data). By moving inactive data to the cool tier, you can reduce the total cost of ownership for your storage. The cost savings come from not having to store all data on high-performance (and more expensive) storage tiers

while there is a trade-off in terms of throughput and potential performance impact, the cost efficiency of enabling cool access depends on your specific workload and data access patterns. If a large portion of your data is cold, the savings from reduced storage costs can outweigh the need for overprovisioning to maintain performance for hot data.

For more detail information, please check below document,
**
https://learn.microsoft.com/en-us/azure/azure-netapp-files/manage-cool-access?tabs=ultra
https://learn.microsoft.com/en-us/azure/azure-netapp-files/azure-netapp-files-service-levels

---

Please do not forget to "Accept the answer” and “up-vote” wherever the information provided helps you, this can be beneficial to other community members.           

If you have any other questions or are still running into more issues, let me know in the "comments" and I would be happy to help you.