What's new in ONTAP 9.8

By | 20. October 2020

Hi and Hello, it’s NetApp INSIGHT time again (albeit virtual) and thus also the time in which new products and features of NetApp are announced.

In this article, I’ll summarize all the new features around the world’s leading storage operating system, ONTAP 9.8.

Improvement in the user interface

Firmware updates via ActiveIQ System Manager

Many of you have certainly waited a long time for this. With ONTAP 9.8, it will now be “finally” possible to perform firmware updates for disks, shelves, service processors, and even the universally “popular” DiskQulification Files via the ActiveIQ System Manager.

ONTAP 9.8 Firmware Update
File System Analysis

There will be a FileSystem analysis directly in the Active IQ System Manager. Capacity breakdown, file and directory number, file activity and age, file usage, etc. you can then display in the CLI without any elaborate rumbling.

ONTAP 9.8 File System Analyze

More from the Data Fabric

What’s new about fabric pool

NetApp has “secretly” integrated the S3 protocol into its ONTAP software since ONTAP 9.6. With ONTAP 9.8, S3 is also visible to you.

FabricPool now also on another ONTAP system

Previously, I was able to use the FabricPool only with StorageGrid or other S3 destinations in the cloud, so it is possible to use a conventional FAS system as a FabricPool target starting with ONTAP 9.8. So if you have a primary ONTAP system and want to put cold blocks or snapshots on a cheaper ONTAP system, you can do so now. (Preset an Approval Process if your tiered Data is greater than 300 TB)

FabricPool now also for HDDs!

Until now, only customers with an All-Flash system or SSD only units could benefit from the FabricPool functionality. This also changes with ONTAP 9.8. Even units with rotating spindles can use the FabricPool functionality to provide cold blocks or snapshots from the primary system.

SnapMirror to an Object Memory (S3)

With ONTAP 9.8, you can also back up to a cloud target at object storage level (S3). Previously, you could only make one SnapMirror into a Cloud Volume ONTAP (or, as has been possible for years, on another on-premises ONTAP system). Now SnapMirror goes to S3. A SnapMirror Cloud license is likely to be required. The licensing is done per TB/ per cluster.

If you’re not a cloud friend, you can also turn a SnapMirror into a local storage grid.

To orchestrate the SnapMirror to S3 relationships, NetApp recommends using Cloud Manager or third-party partners such as VEEAM or CommVault. A Native NetApp orchestration is to be added soon. A restore only works when I turn the SnapMirro over again. So I can’t take the volume online in the S3 bucket.

News from FlexGroups

A lot has happened at FlexGroups. I’m just trying to accommodate this one in a puncture point:

  • Datastore Support for VMware
  • Support for the Virtual Storage Console
  • SnapCenter Backup Support
  • There are now also 1023 SnapShots supported in the FlexGrouup
  • NDMP backups can now be restarted

Simpler Tech Refreshes

To simplify the regular tech refreshes, you can now

  • add the new systems to an existing cluster,
  • the version difference of the ONTAP software can be up to 2 years (from ONTAP 9.8),
  • Use Volume Move to migrate the data to the new hardware,
  • and then remove the old nodes from the cluster
Simpler Tech Refresh

SnapLock Updates

SnapLock, the WORM function of ONTAP, now also supports a volume move. This allows you to move compliance protected volumes to another aggregate without having to make a second copy.

MetroCluster Updates

Hallejulia! Unmirrored aggregates are now officially supported! From now on, you can only protect really business-critical data with the MetroCluster. Data that is not so “valuable” for you can only be on one side of the Metrocluster.

Transition from 4-node Fabric MetroCluster to 4-node IP MetroCluster

For those who want to turn a tech refresh of their old 4-node MetroCluster over Fabric MCC-FC into a 4-node MetroCluster over IP MCC-IP, there is now the possibility to do this without downtime. The 4-node MCC-IP is added to the MCC-FC during the transition and the data is migrated via volume move. (See above Simpler Techrefreshes). A transition of a 2-node MCC-FC still requires a downtime.

Phew… again quite a lot and still far from complete. Persistent ports for All SAN arrays, larger LUNs up to 128TB (ASA only), FlexCache SMB support, QoS for Qtrees, and much more…

I’m trying to keep you informed of what’s going to be released.

Best regards


DISCLAIMER: This post represents my personal observations and is not official from NetApp or other authorized ones. misinterpretations or misunderstandings.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.