MSPs (Managed Service Providers) are companies that handle a client’s IT infrastructure remotely. The arrangement between the two parties (Client and the MSPs) are on a subscription model wherein the client pays MSP based on use.
MSPs typically handle both IaaS (Infrastructure as a Service) and PaaS (Platform as a Service) implementation for their clients.
Clients benefit immensely by opting for MSPs to solutionize and implement their applications on the cloud effectively.
These days, MSPs have a huge responsibility to play, as most of the technology innovators are shifting towards cloud implementation of both their existing traditional applications and also developing completely new ones in the cloud. According to a case study by Gartner, it is estimated that by 2024, nearly 50% of the deals involving cloud service will comprise application development services and cloud infrastructure professional and managed services. This is a meteoric rise from just 10% in 2019.
Why are MSPs needed?
Even though IaaS and PaaS are automated and self-serviceable, the requirement of IT operations in managing and implementing them cannot be ruled out.
Clients have to correctly choose the right service elements and configure them properly. The same clients also have the added responsibility of maintaining those aspects of the service they possess control of. Some of these responsibilities are configuring guest OS, configuring middleware and applications running on their IaaS compute instances, etc.
Another challenge is that the latest types of platform services usually require constantly evolving skills and management expertise. Challenges in governance, finances are also common when implementing IaaS and PaaS.
Owing to these, clients constantly lookout for third-party solution providers or the MSPs to assist them. A client engages the MSPs on a short-term basis which in turn is a part of a long term infrastructure management strategy.
Two use cases of cloud adoption
The cloud adoption of IaaS and PaaS predominantly consists of two use cases. They are –
1) Rehosting existing applications that were developed using traditional architectural precepts into the cloud
2) Creating new or completely replacing a cloud-native architectural precept.
Earlier, both the above use cases used to be independent as clients were implementing either one of these and rarely both. However, in recent times, clients have started to pursue both the use cases in parallel. They not only migrate an existing application to the cloud from their on-premises data-center but also begin a long-term creation and transformation of applications to modernize them and absorb the benefits of wholly creating an application in the cloud.
MSPs need to handle both the use cases even if they specialize in either one of them. They also need to have high levels of technical expertise in working with top cloud providers such as Google Cloud, AWS, and Microsoft Azure, etc. to successfully implement these use cases.
Typically they are expected to have the capability to provide cloud-optimized solutions irrespective of which use case they are handling with a client.
Bimodal nature of Cloud Adoption
Cloud Adoption is bimodal which means they manage two separate but coherent styles of work. One of the modes(Mode 1) has a prime focus on predictability and safety. While the other mode(Mode 2) focuses on exploration and agility. However, even these two modes have evolved over the years and the way they relate to cloud adoption today is not what it used to be years back.
Earlier Mode 1 was mainly used to attribute to the rehosting of applications built in traditional non-cloud architecture. Since such applications were considered to part of a ‘stable IT’, they tend to be associated with Mode 1 owing to its focus on predictability and safety. On the other hand, Mode 2 was synonymous with cloud-native applications which are considered to be more dynamic and experimental which is in line with the focus of Mode 2 – exploration and agility.
Today, both modes are equally relevant to cloud-native precepts and have also lost their relevancy to evaluate an MSP. Instead, MSPs are evaluated based on the two use cases of MSPs discussed earlier.
Functions of an MSP
An MSP, when chosen by a client, is expected to perform the following functions for a successful cloud implementation –
- Design solutions that utilize cloud service providers like AWS, Microsoft Azure or GCP in the best possible way and add value while migrating the existing application or while building new cloud-native applications.
- Completely utilize the breadth of a Cloud Service platform by employing not just the IaaS components of it, but also the PaaS components to provide the maximum output.
- Utilize the best practices of cloud infrastructure management such as using immutable infrastructure and DevSecOps wherever possible
- Leverage the native managed capabilities of the cloud service provider and use it to automate services. Use DevOps tool and other automation tools to leverage ‘infrastructure as code’
- Have a management portal that provides cloud service expense management and also helps in the governance of the cloud provider accounts including the integration with identity management.
- Provide technical support to various roles in the client’s organization such as infrastructure and operations (I&O) technicians, application developers, etc.
- Optionally handle the details of the cloud implementation that are not directly related to its implementation. For example, in virtual machines, an MSP may handle the patching of the OS and also its configuration, integration of the middleware and also manage any details that are specific to the database and application integration.
- Provide supplementary service for components that are not part of the cloud provider. Hybrid IT which is a traditional non-cloud infrastructure that is managed along with the cloud is one such example wherein the MSP needs to provide services to the non-cloud infrastructure also.
MSPs have a huge role to play and have a host of functions that need to perform when chosen for a cloud implementation by a client. Gone are the days when an MSP had to only rehost an existing application from a non-cloud infrastructure to cloud, or only create a new application complete in the cloud. These days, MSPs have to handle both these use cases together. Since MSPs have the scale, breadth and depth of experience and can provide 24X7 DevOps, Support, SRE (Site Reliability Engineering) and deliver them at a cost effective economic package, prospective organizations should consider MSPs as a part of their Cloud Strategies.