In today’s digital era, storage is an essential element that can make or break the customer experience. Since cloud has become an essential pillar for the digitalization of enterprises, software as a service (SaaS) plays an integral role in the overall storage transformation too. SaaS is commonly known as Software-as-a-Service, but when we talk in the context of storage offerings on the cloud market place, it refers to the software edition of enterprise storage available on hyper-scale cloud providers’ platform.
There have always been questions whether it is beneficial to adopt SaaS storage and backup services from the public cloud marketplace when similar offerings are already available in the public cloud portfolio. For instance, why should the customer go for Dell/NetApp/Pure Storage/IBM storage and backup SaaS-based version when all these services are readily available on the public cloud with nomenclature S3/blob/EBS. Here, we would try to demystify some of these myths and answer some of the relevant questions generally asked in this regard.
The answer to the above queries depends on two factors: the first is the customer use case and the second is the overall ROI benefits that the customer will get in the long run along with technology improvements in the existing services model.
Therefore, let's discuss the first factor wherein the customer has a 3-tier architectural environment where the database and code applications are deployed on-premises and the web applications are hosted on a public cloud. This is quite a standard model of application deployment that has been there for decades. However, if we think about the challenges in this model, then we will find certain pitfalls concerning the performance, security, compliance as well as lack of long-term ROI benefits. These challenges can be mitigated by adopting a hyper-scale cloud marketplace storage strategy.
So, let us presume that the customer’s front-end web applications are hosted on AWS EC2 instance along with secure connectivity (direct connect) towards on-premises database. On certain days when volumetric transactions are at their peak, end-users face a possible transaction delay. Though, at the backend, there are multiple DB sessions established with web applications, the data refresh rate still takes significantly more time to commit the record in the on-premises database.
Apart, from the performance challenge, there were two more gaps identified in the above-given model which also impacts the TCO and these are high egress costs along with consistent link charges towards public cloud either from the co-location or customer DC. Coming back to the initial point whether it's beneficial to lift and shift all the applications to the public cloud or use a SaaS-based model of storage and backup services?
So, for understanding, let us examine two scenarios. In the first example, the challenges were identified when the database and the application moved to the public cloud (AWS), and the next example is how to mitigate the problem by adopting a SaaS-based storage model.
Do we think that moving the applications and the database to the public cloud will resolve the performance issues? Though, in the above solution, we have hosted both applications (web/DB) on the same platform, unfortunately, the performance issue still persists. The reason behind the performance lag is when applications directly access cloud storage it decreases the maximum potential performance of storage due to shared network, connecting EBS and EC2. In such cases, the random read and sequential write performance get impacted.
Therefore, how can this problem get addressed through a SaaS storage service model?
Let’s take a similar scenario with one of the storage vendors NetApp SaaS-based cloud solution, NetApp Cloud Volumes.
Just for quick reference, the +1NetApp cloud storage is a SaaS model that runs on a virtually deployed machine on a public cloud. The underneath storage either comes from AWS EBS, Azure manage disk, or GCP storage.
Coming back to the point where we are discussing how to mitigate read and write latency issues with the NetApp SaaS storage model.
The below illustration shows that applications are hosted on the OS where the disk is provisioned from cloud storage Ontap, which is Flash cache-enabled and allows recently used data to get cached on local attached NVME storage. Now, all the random reads will be served from the cache instead of hitting underlying disks (EBS). In this way, the application which requires more random read tasks like mail/database/file servers gets significantly better read IOPS and boost performance.
Cloud Volumes (CVO) over EBS
Apart from performance improvement, another advantage of adopting the SaaS storage model is to make a complete ecosystem of on-prem and public cloud when integrated with public cloud services. This will provide benefits like space-saving, easy deployment of clone, high availability, and others.
For industries like pharmaceuticals and automobiles, which always strive for quick production-data clones for R&D on the public cloud within their geographical boundaries, HCL can play a vital role by developing a framework with the amalgamation of on-prem and SaaS-based storage and protection solutions.
A few months back, Pure Storage launched its GA version of +2Cloud Block Storage (CBS) for AWS. CBS is a SaaS version of the Pure Storage on-premises AFA model. CBS also has the underlying storage from EBS and on top of it deploys its own OS name purity, which has many rich storage-saving features as compared to EBS as well as high availability features. Pure Storage claims about more than USD 8 million saving over 5 years with the adoption of the SaaS model (CBS) on the public cloud.
Dell also has shown an extremely rich SaaS-based solution on the public cloud market place especially on AWS. +3DDVE is an Dell distinguish SaaS-based data protection solution. DDVE helps customers to minimize the TCO by integration with AWS S3. Also, the customer can use BYOL while deploying DDVE on a public cloud. There are many other features with DDVE that help in minimizing the overall TCO for data protection.
Traditionally, we have always trusted hyper-scale public cloud services for minimizing TCO and better ROI. Up to some extent, this is true, but when we deeply analyze customer's use cases, it may not be a good approach to always lift and shift every service towards a public cloud. Especially with respect to storage and backup solutions directly associated with performance, capacity optimization and compliance. Hence, it's not mandatory that the public cloud services fit all scenarios.
In recent years, the HCL engineering team has been doing extensive research along with its key OEM partners for the adoption of a SaaS-based model of storage and backup services on a public cloud. HCL can play a significant role in developing a holistic ecosystem with SaaS-based storage and backup solution for customers who are bound with data governance and compliances.