#FSXn aggregate best practice

1 messages · Page 1 of 1 (latest)

shell flax
#

Hopeing someone could clarify this for me.

Typically on most Storage systems best practice is to keep utilization below 80%. Is this also true for AWS FSXn?
If so what are the reasons? Also is this documented anywhere?

Would the following statement be true?
"For AWS FSx for ONTAP, best practice for aggregate capacity involves maintaining an average utilization below 80% on an ongoing basis, ensuring sufficient space for new data and maintenance operations, and potentially rebalancing workloads if aggregates become imbalanced"

sinful parrot
#

There is no such best practices in term of performance, but it’s recommended to keep enough space to allow enough time for the SSD pool to expand. The space will be available very fast, but the background processing can take multiple hours and in that time you won’t be able to grow again

#

Rebalancing is more relevant for scale out filesystems with FlexGroups

shell flax
sinful parrot
#

Yes, for scale out for cases when you are adding HA pairs to the filesystems

#

It’s less relevant for scale up

#

Workload factory BTW has a nice feature to rebalance on the block level if rebalancing is relevant for your use case

shell flax
#

I really would like to use WL factory but we don't have it in our environment. We often run over 80% on many of our file systems. Sometimes , we go over 95% but then when we try to expand the pool it can take many hours. Currently we look at expanding when the pool gets to 85%

sinful parrot
#

It’s a free service. We can help you onboard and run an overview

#

This can help you expand the capacity when reaching a threshold

vapid leaf