According to industry research, the hybrid cloud market is estimated to grow from $33.28 billion in 2016 to $91.74 billion by 2021, at a compound annual growth rate (CAGR) of 22.5 per cent. What industry research such as this doesn’t explain, however, is that:
- The hybrid-cloud picture is far from clear, due in part to the lack of agreement as to what hybrid cloud actually is. And of course it’s hard to work with what isn’t well understood. Theoretical physicist Richard Feynman is famously quoted as saying, “If you think you understand quantum physics, you don’t understand quantum physics.” To this extent, some would argue that hybrid cloud is a little like quantum physics.
- Organisations struggle to get hybrid cloud right at their first attempt, partially due to the nature of hybrid cloud, and partially due to common hybrid cloud adoption errors.
Why organisations struggle to get hybrid cloud right at their first attempt
There are a number of aspects of hybrid cloud operations that corporate IT organisations fail to recognise and address, including that:
- Hybrid cloud is complex, takes a long time to get right, and can be very expensive. You can’t just buy a hybrid cloud product, and there’s no vendor with a single hybrid cloud SKU. It’s always a custom-built solution with all the issues that this type of IT project usually experiences, especially in terms of under-estimating complexity, over-running on cost, and not meeting cloud-consumer needs.
- There’s a lack of hybrid cloud practitioners. A broad skillset is required to deploy a hybrid cloud. Beyond the myriad of technologies to integrate, both on-premise and in the public cloud, there’s the required integration task to “bind” the clouds to make them hybrid. The biggest challenge – and the main cause of private cloud failure – is modifying the organisation’s operational model.
- Cloud appetites are potentially already moving away from hybrid cloud. Application owners and application developers are moving up the stack – stepping away from being in the cloud infrastructure game. Application owners and consumers, such as CRM application users, are moving to software-as-a-service (SaaS) models. Application developers are moving to platform-as-a-service models (PaaS) – these might run on a private cloud but they are more at home in public cloud and there’s usually no requirement for hybrid cloud.
With the world of hybrid cloud fraught with complications and tensions even before a company first dips its toes into the waters, there’s a lot to consider.
10 things to avoid when starting with hybrid cloud and their workarounds
A company’s first foray into hybrid cloud is definitely an opportunity to learn from others. It’s an opportunity to consider, and to avoid, the wide range of potential problems and pitfalls that other companies have unfortunately experienced as they started with hybrid cloud. These include:
1. Overlooking security and compliance
This isn’t just important for organisations that need to comply with industry requirements such as the US Health Insurance Portability and Accountability Act of 1996 (HIPAA) or the Payment Card Industry Data Security Standard (PCI DSS). Security and compliance needs to be front-and-centre for all organisations, with them needing to determine where to store sensitive data, how to protect data in transit, and how to encrypt and secure data at rest.
2. Underestimating the people requirements
As mentioned earlier, when introducing new technologies, organisations can easily forget the total skillset required. And, in the case of hybrid cloud, even if the full skillset has been identified, organisations might still struggle to acquire the right people in a very competitive marketplace.
3. Understating the required data centre footprint and costs
Besides allocating the right local footprint, and leaving room for growth, organisations also need to take into account, and budget for, the different investments in operational costs associated with co-locating services and the capital costs associated with purchasing hardware for local operations.
4. Not fully planning for disaster recovery
Building in redundancy and robustness is an essential part of assuring availability for any IT environment. So don’t forget to make the right decisions to achieve the required RTOs (recovery time objectives) and RPOs (recovery point objectives) for the service or services supported by the hybrid cloud environment.
5. Overlooking the need for seamless interoperability and compatibility
With hybrid cloud, organisations need to “stretch” their in-house environment to the cloud, and it’s therefore important that they ensure both environments are compatible with each other.
6. Underestimating network performance and bandwidth needs
In addition to having both the in-house and cloud network configurations the same for easier management, organisations must take bandwidth considerations into account. It’s especially important to calculate the amount of traffic being transferred, moved, and replicated, and the bandwidth and latency required to ensure that the networks can manage the load.
7. Insufficient effort placed on finding the right vendor
Whether it’s compliance with different standards, physical location, security requirements, scalability, flexibility, or any of the non-technical aspects of the vendor – all these and more need to be taken into account when choosing the right vendor, not just the right technology, for the job.
8. Not talking with other hybrid cloud customers about their failures and successes
Meet and speak with other organisations, with similar needs to your own, to find out the key learnings from their hybrid cloud experiences. Learning from your mistakes is good, but learning from the mistakes of others is even better.
9. Failing to start small
Start with a trial. As with any big change, it’s recommended to trial things first and to build up operations in increments. In this way, your operations and the required changes to it can be evaluated in real-time with adjustments made as needed.
10. Lacking procedures and documentation
For some, having a high level of operational formality might be mandatory for compliance purposes. But, regardless of industry-specific regulations, having recognised procedures and documentation will help from the planning stage, through transition, and into business-as-usual operations.
Sarah Lahav, CEO of SysAid Technologies
Image Credit: phloxii / Shutterstock