Storage
Mobility
Protection
Analysis and control
TOPICShare
This week I was invited to join the Tech ONTAP Podcast to talk about a brand-new feature for Cloud Volumes ONTAP: FlexCache®.
I sat with Justin Parisi, the host of the show, as well as Shriya Paramkusam, the product manager for FlexCache and FlexGroup, and Chris Hurley, the technical marketing engineer for FlexCache and NAS storage at NetApp (you can find him on Twitter as @averageguyx). It was a lot of fun, so please have a listen below!
What is FlexCache?
FlexCache in NetApp ONTAP is a caching technology that enables sparse writable replicas of volumes on the same or different ONTAP clusters. It can bring data and files closer to you for faster throughput with a smaller footprint.
On the podcast we talked about FlexCache, and more specifically, using FlexCache with Cloud Volumes ONTAP. FlexCache has been around for a while but it’s only with the latest ONTAP version, 9.6, that it’s coming to the cloud (along with a bunch of other brand-new features).
FlexCache is a replicated volume that lets you extend your volume namespace out beyond the confines of your data center, clusters, or SVMs. This can help increase compute power, keep latencies low, and get the best network connectivity possible. FlexCache allows you to replicate parts of the data to a target Cloud Volumes ONTAP instance in a remote location so users anywhere can collaborate with no performance lag. And the source and target copies are kept in constant sync so the data each user is working with is always current. This is suitable for distributed working environments.
Similar to FlexClone cloning or using SnapMirror, FlexCache creates copies of a volume based on WAFL technology. This FlexCache copy will cache your reads locally to any clients connected to it. One big advantage of FlexCache replicas is that you can extend an on-prem installation with all its data and networking directly into the cloud without any of the heavy lifting of a full migration.
There’s a lot more to get to know, so please check out the full podcast here:
This is the first time that FlexCache is bringing its abilities to consolidate disparate data centers to the cloud and the biggest impact it might have is in providing a more-performant way to hybridize your workloads.New for the Cloud
Justin brought up a good point in the discussion about this: If you’re in a hybrid architecture, and you’re working on an intensive workload that needs a lot of compute (his example was an EDA environment) the workload and the CPU need to be as close as they can get. If the compute is on-prem and the data is in the cloud, there’s a lot of distance. FlexCache erases that distance by caching reads to the closest client, making it faster for the storage and CPU to communicate. Performance stays at peak levels even across a complex environment.
A capability like this also helps meet increasing demand spikes that happen in file sharing, adding one more advantage that Cloud Volumes ONTAP has in meeting file sharing challenges that the managed file services just can’t match.
Plus, the FlexCache volume is completely writable, so anything that changes in the cloud can easily be reflected back at the source. This is something we term “write-around” since the changes aren’t directly written to the FlexCache copy. There’s encryption benefit, and, as Shriya mentions, a huge number of cache volumes per node that systems can host (now up to 100 from 10 in the previous build). There’s also a hugely improved networking protection to make sure that if there is any kind of network failure and the cache is disconnected, the reads in the cache are disconnected and the writes will be done directly at the source. When the network is back up, FlexCache does an efficient resync process comparing any changes that don’t exist in the cache and then updates them.
Add it to the existing list of Cloud Volumes ONTAP benefits, including:
- Storage efficiencies to keep cloud storage costs from getting out of control.
- Snapshots that are a faster way to back up your data and which serve as the foundation for seamless DR process.
- FlexClone instant cloning to speed up the dev/test and release cycle.
- Zero-data loss high availability to keep running when instances fail.
- And easy data migration capabilities to get you to the cloud faster.
