Trekintech

Management in ONTAP 9

With the release of ONTAP 9 I wanted to do a short overview of the new management capabilities. Over the past few releases simplicity has been a core driver of ONTAP releases and it's great to see this continue in the latest release whilst still providing new capabilities.

System Manager

The first thing that anyone who has upgraded to ONTAP 9 will notice is that System Manager has undergone a design overhaul. Gone are the tree style menus that made you feel like you were labouring towards the object you wished to interact with. In it's place is a much more modern and logical design that makes management simple. I'm a big fan of the global search function which allows admins to quickly get to the object they need without having to cycle through menus.

System Manager also provides much greater visibility into performance than prior iterations , with ONTAP 9 you have the ability to drill down on performance through each element with a view of latency, IOPS and throughput available on all storage objects. Helpfully related objects are also hyperlinked allowing you to drill down into nested entities.

One of the biggest changes is the introduction of Application provisioning wizards which create all the necessary storage elements for applications based on NetApp and application vendor best practices. This is available with the AFF currently , hopefully this will make it's way to all variants of hardware in the coming releases.

Additionally System Manager provides support for Snaplock volumes, RAID TEC and the ability to generate an AutoSupport.

Unified Manager & Performance Manager

For those that aren't aware Performance Manager is now embedded within Unified Manager in that you no longer need to sign on to two different portals. Performance events hyperlink directly into Performance Manager as of version 2.1.

The initial dashboard provides an overview of the cluster with the ability to drill into Health or Performance Details

There is also a section for your favourite storage objects which provides a overview of selected objects without having to drill down.

The main changes from Unified Manager come from the Performance Manager component however Unified Manager itself does now support EMS events which once captured support the ability to execute custom scripts. In order to see the full list of EMS events in ONTAP 9.0 use event catalog show in the cluster shell.

Other than that Unified Manager continues to provide a simple management interface into monitoring and alerting at scale. In built reporting also allows scheduling of the inbuilt or custom designed reports in a variety of formats so that you can keep a close eye on storage growth and snapmirror health. Also be sure to check out the automation store for community reports aswell as workflows to add to Workflow Automation which I'll cover separately.

The big feature to make it into the latest release of Performance Manager and ONTAP 9 is the performance headroom functionality. This feature uses analytics to understand how much headroom is left in your storage array before latency is impacted therefore ensuring accurate performance planning even with an ever changing workload. The dashboard now shows the performance % used by the current workload and comes complete with failover planning so that you can understand the impact of a node failover.

Performance Manager also shows Top Performers for each object and allows the ability to overlap objects to understand the impact of each objects performance on that of another and carries 13 months worth of history.

Performance Manager also has the ability to define performance thresholds which when breached for the specified duration creates an alert which will also provide victim and bully analysis along with suggested fixes. Dynamic thresholds are also used, the system understands the normal performance metrics and advises when there is a deviation from the norm. Additionally the system learns from your response so should you keep ignoring what is deemed a performance breach the system will stop alerting you and will consider it a false alarm.

Overall the management, monitoring and reporting has come on leaps and bounds in the last few releases and adds some great functionality aswell as simplicity improvements in ONTAP 9. It would be great to hear feedback from customers and partners on the new versions and where else the software could be improved.

Author image
About Mike Andrews
Manchester, England Website
Senior System Engineer currently working at Cohesity focussed on Data Management. I have a keen interest in cloud and have worked in the IT industry for 13 years. My views are my own.