![](https://static.wixstatic.com/media/981170_aac6b6730d9e419ebb653a182d881fcb~mv2.jpg/v1/fill/w_980,h_433,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/981170_aac6b6730d9e419ebb653a182d881fcb~mv2.jpg)
These days the scope of cloud platforms has extended beyond just one provider. With increasing demand on data protection and data confidentiality to meet regulatory compliance, the boundary of underlying infrastructure has grown to a combination of multi cloud and on-premise models.
So what really drives an organisation towards a hybrid cloud model and how must it be designed to achieve those goals.
Here are the top 10 Driving and Design factors for any Hybrid cloud setup:
5 Most Common Factors Why Hybrid Cloud Is Chosen:
Data localisation/residency laws in many countries nowadays, requires organisations to host PII (Personal Identifiable Information) data within the same country. Such regulations often force business to maintain their data components within an on-prem data center while using a cloud service or a SaaS product for their their application layer.
Protection of Intellectual Property is another reason which can drive adoption of Hybrid clouds faster. For instance, some cloud providers might be sharing your data within their own network but it different regions than yours. This is often true in cases of managed cloud services like monitoring tools, vulnerability scanners where the managed console is hosted in a shared network. This might not comply with the data privacy , confidentiality or sharing policies of your organisation.
RTO and RPO are the factors which will eventually measure your service/business SLA. It will decide how fast you can recover (RTO) and how much can you recover (RPO). To achieve the desired numbers a DC/DRC setup is a must. Now almost all cloud providers promise to deliver H/A , but very few actually comply with a DC/DRC. And even if they do, it may not really deliver the RTO/RPOs. You will only realise this caveat once you do a DR Drill with the Cloud Availability Zones/Regions which is almost impossible. Such limitations also drive towards a hybrid cloud setup for mission critical businesses.
Reduce Truck Factor, a.k.a Don't put all eggs in the same basket. While the top three reasons can drive more towards an (on-prem + cloud) hybrid model , this one in particular is one of the reasons to drive a Cloud A + Cloud B kind of hybrid model. This helps create a redundant/alternat network for always on services and helps to mitigate vendor-lockin issues.
Service and Support Costs, last but not the least, is also one of the reasons why organisations choose to maintain a hybrid cloud model. While some providers can be very cost effective with their discounted commitment models but their services might not be as per your needs. On the other hand, there can be provider with highly efficient and performant services with higher costs. In such case a mix and match of both can be cost effective.
Although, all 5 are vary valid reasons and a Hybrid cloud model may seem helpful but it can also turn to a disaster if not designed correctly. Infact to achieve the desired outcomes for all the driving factors mentioned above there are certain designs that every hybrid cloud models must have.
5 Must Have Configurations For Any Hybrid Cloud Setup:
1. Dedicated Interconnected Network:
For any hybrid cloud to succeed it must have dedicated links or private links connecting its sites. This will ensure :
Security of data in transit
Minimum latency for data transfer
Better performance for service to site communication.
Fault tolerance between sites.
Multi Site Data Replication for DR
2. Single Secure Entrypoint:
There must be single endpoint that provides secured access to all services from the internet. This entry-point must cone enabled with:
DNS based backend mapping
Security Controls like DDoS Protection
Single Trusted Point Endpoint exposure
3. Modular Infrastructure as Code:
Select tools that can wither be agnostic or support multiple providers
Segregate Source modules/playbook/cookbook from Provisioners
Always do test driven delivery of infrastructure.
Create environments on demand based on utilisation.
4. Centralised Command Centre:
Centralised Cloud Administration
Prevents Vendor Lock-in
Better Observability
5. Self Managed Stateful Services:
IP(Intellectual Property) Protection
Data privacy and Confidentiality
Hidden Cost
Vendor lock-in factors
With all the factors combined together heres how a hybrid cloud value stream must look like:
Hybrid Cloud Value Stream Map:
![Hybrid Cloud Factors](https://static.wixstatic.com/media/981170_a1851105585a4231a311b93e8029a71e~mv2.jpg/v1/fill/w_980,h_603,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/981170_a1851105585a4231a311b93e8029a71e~mv2.jpg)
If you like this article, I am sure you will find 10-Factor Infrastructure even more useful. It compiles all these tried and tested methodologies, design patterns & best practices into a complete framework for building secure, scalable and resilient modern infrastructure.
Don’t let your best-selling product suffer due to an unstable, vulnerable & mutable infrastructure.
Thanks & Regards
Kamalika Majumder
Comments