Skip to main content

Serverless for dummies


One of my first blogs was an introduction to non-technical people about Docker (https://www.linkedin.com/pulse/docker-dummies-joris-lochy) and last year I wrote a similar blog on DevOps (https://www.linkedin.com/pulse/devops-dummies-joris-lochy). Recently I joined the "Merit-VC" platform (i.e. https://www.merit-vc.com/ - a platform to train yourself in VC-funding), where I got the assignment to write a report on the trend of "Serverless". I thought it would be interesting to share my conclusions in the form of a third "for dummies" blog, introducing another innovative technical concept to non-technical people, namely the concept of "serverless".

Before introducing serverless it is however important to have a notion about cloud computing. Cloud computing allows to "rent" servers from cloud companies instead of having to setup and maintain your own infrastructure (data centre). The advantages are a switch from CapEx to OpEx (you pay only for what you use), a much higher flexibility to onboard/offboard servers (scalability based on your business needs) and you can avoid having to attract the specific skills required to properly operate infrastructure. Additionally most cloud providers have built all kind of abstraction layers, which allow to setup and maintain servers and the software deployed on them easier and faster.

This cloud market has known an exponential growth in the last years, which was even further accelerated by the Covid-crisis. Based on a report of Canalys, the market was estimated to around $142 billion for the year 2020. In the market there are 4 dominant players, i.e. Amazon AWS (31% market share), Microsoft Azure (20% market share), Google GCP (7% market share) and Alibaba Cloud (6% market share, although this is mainly in Asia). This means also that there are dozens of other parties with less than 5% market share still forming together a total market share of 35% (like Oracle, IBM, Tencent, DigitalOcean, SalesForce, Verizon Cloud…​).

The most common service bought from a cloud provider are still servers (which are actually virtual machines at the cloud provider), like Amazon EC2, Azure Virtual Machines, Google Compute Engine or DigitalOcean Droplets. These servers can be added/removed via a simple user interface (with the user being able to choose the size of the virtual machine, the availability…​), giving a very dynamic server management. On the server, the company can then install any software.

Over the years cloud providers have however started to offer more and more services, which abstract away the underlying infrastructure. The most common ones being:

  • Storage services, i.e. a simple service where files can be stored and retrieved via an easily accessible API. Most known examples are Amazon S3, Azure Blob Storage, Google Cloud Storage, DigitalOcean Spaces…​

  • Database services in all forms, like the most common SQL databases (like PostgreSQL, Oracle, MySQL or Maria DB) and noSQL databases (like MongoDB, Cassandra, ElasticSearch, HBase…​). Typical examples being Amazon RDS, Amazon Dynamo DB, Azure SQL Database, Azure Cosmos DB, Google Cloud SQL, Google Cloud BigTable, MongoDB Atlas…​.

  • Docker container management, like running containers in a flexible, scalable and reliable way, all or not on Kubernetes. E.g. Amazon ECS, Amazon EKS, Google GKE, Azure AKS, Red Hat OpenShift Container Platform…​

  • Content management systems and eCommerce platforms, like Wordpress, Drupal, Magento…​

  • Identity management, like Auth0, Amazon Cognito, Azure AD…​

But this is just the tip of the iceberg. Every cloud provider has an offering of hundreds of managed services, making it sometimes extremely difficult to find your way in this offering.

These abstractions have led to the new term of serverless. This doesn’t mean that there are no servers anymore, but rather that the underlying servers are fully abstracted away for the consumer of the service, i.e. the allocation of the infrastructure (server) is fully managed by the cloud provider.
The services can be the services described above (also called managed services, managed cloud, backend-as-a-service or BaaS), but it can also be services directly offered to developers in the form of functions which can be requested to be executed via an API. This type of serverless is called "Functions-as-a-service" (FaaS), with the most known ambassador being AWS Lambda, which is the implementation of Amazon of this service. Of course the other cloud providers have in the meantime similar offers like Google Cloud Functions, Azure Functions or IBM OpenWhisk.

Simplified, this type of serverless means the consuming company requests to execute a specific function (a piece of code). The cloud provider then deploys the software component (at run-time) on one of their servers (totally abstracted away for the consumer) and executes it. Theoretically this means that a company does not have the deal anymore with servers, has almost unlimited (auto-) scalability and only pays for what it consumes (i.e. you never pay for idle resources). The downside is the complexity and the performance. The complexity comes mainly from the fact that applications become a complex orchestration of small units (most cloud providers set quite strict time-out times to execute a FaaS requiring software to be split in small units) which are executed and from the fact that each function needs to be fully stateless. This means any data (used as input or generated as output by the function) and logging/tracing information needs to be stored by a separate service. The performance impact comes mainly from the time needed to deploy the function on a server. This is a domain where cloud providers try to optimize, by keeping certain functions "hot" (already installed) on a server and by optimally predicting (and pre-installing) which functions will (soon) be required (in the near future). However these techniques have their limits, as obviously a cloud provider cannot keep all functions installed on several server instances (to cope with potential loads), as this would make the economic model for the cloud provider impossible.

The ultimate goal of both serverless techniques (i.e. BaaS & FaaS) is of course to save money for a company. While in many cases these techniques can save money, this is not always the case. As always, it depends on the specific needs of the customer. Typically if you have a high and consistent (not a variable) usage and are a large corporate, these solutions might not be the best option (as price will likely be higher and the increased complexity and potential performance impacts outweigh the advantages). However for other use cases, like e.g. a start-up or scale-up, with very bursty and changing loads, and with a limited team having to cope with all IT and infrastructure aspects, this kind of services can be ideal.

However next to the infrastructure and scalability impacts, it is also important to know that FaaS is also a methodology, i.e. a way to implement software, with following characteristics:

  • Code is split into small modular components (microservices), which are orchestrated together for the software

  • Every component needs to be stateless and its execution should be finished in a limited time. The result is that such a component is by definition fault tolerant, as the function can be relaunched without any specific action in case of issues

The gains that can potentially come (in the long-term) of this development methodology can outweigh the possible infrastructure cost gain. In most companies, although infrastructure is a big cost, it is still marginal compared to the human cost of IT people working in IT change and in run departments.
As such the following questions should also be raised:

  • How much time can I gain in the IT teams with this new way of implementing software?

  • How easily can people collaborate together in this architecture?

  • How easy is it to onboard additional people?

  • How sure I am, I will not have to refactor everything, because the associated technology has become obsolete?

  • How easy can I find IT people mastering the underlying technology?

  • …​

The division of software applications in such smaller components is one way to give a more positive answer to those questions, but it depends of course on the type of software, the maturity level and culture of the organization and the character and style of the IT resources.

Independent of those reflections, it is clear that serverless is here to stay and will likely increase in the coming years.
As such following trend are clearly visible in the market:

  • Managed services offered by cloud providers become more and more powerful and functional (business oriented). While initially limited to storage and database services, the services are becoming much more business oriented, making the line with Software-as-a-Service (SaaS) and Business-as-a-Service (BaaS) more and more blurry.

  • Cloud providers build more and more abstraction layers on top of technical components. This makes the consumption of those services easier, but it also means that the setup becomes more opinionated and less flexible (cfr. my blog "Abstraction in Financial IT - How far can and should we go? - https://bankloch.blogspot.com/2020/02/abstraction-in-financial-it-how-far-can.html)

  • With software becoming more open source, cloud providers can provide excellent tooling, with very little investment. This creates a strong tension between cloud providers (like Amazon) on the one-side and (open-source) software companies (like MongoDB, Elastic, Confluent or Redis Labs) on the other side. A new form of alliance between both types of companies will be essential, with cloud providers bringing the infrastructure and DevOps tooling and the software providers providing the in-depth expertise of the specific software. Till now many software companies have tried to fight the cloud providers by changing their license policies (cfr. story of Redis), but this makes their product at the same time less future-proof and less desirable.

  • Functions as a Service will become more common, as companies switch more to the cloud, microservice based architecture become more popular, developers become more acquainted with this technology, the prices of FaaS become more competitive and the potential performance issues linked to FaaS (to deploy a function) become better managed.

  • After a first wave of companies shifting from on-premise infrastructure to the cloud, now you see a 2nd wave to multi-cloud setups. This gives the advantage of not being locked-in by one cloud provider, being able to offer a higher worldwide availability and better performance (as none of the cloud providers have data centres in every country) and allowing to choose the best managed service of every cloud provider. E.g. while AWS has the richest and most customer-centric offering, GCP has probably the best offering when it comes to Big Data and Machine Learning processing, while Azure has a lot of advantages via its link with the productivity tools (like Office 365, Teams, Sharepoint…​). However in order to be able to able to work multi-cloud it is important to use a little as possible specific abstractions of cloud providers. This means for FaaS that the specific cloud provider frameworks (like AWS Lambda) should be avoided and open source frameworks like Knative, OpenWhisk, Kubeless or OpenFaaS should be adopted.

Even though enormous steps have been made in serverless in the last years, it is clear this domain is still in its embryonic stage, meaning significant developments can be expected in the coming years. I am looking forward to further following the exciting innovations. 

Labels:

Comments

Post a Comment

Popular posts from this blog

Transforming the insurance sector to an Open API Ecosystem

1. Introduction "Open" has recently become a new buzzword in the financial services industry, i.e.   open data, open APIs, Open Banking, Open Insurance …​, but what does this new buzzword really mean? "Open" refers to the capability of companies to expose their services to the outside world, so that   external partners or even competitors   can use these services to bring added value to their customers. This trend is made possible by the technological evolution of   open APIs (Application Programming Interfaces), which are the   digital ports making this communication possible. Together companies, interconnected through open APIs, form a true   API ecosystem , offering best-of-breed customer experience, by combining the digital services offered by multiple companies. In the   technology sector   this evolution has been ongoing for multiple years (think about the travelling sector, allowing you to book any hotel online). An excellent example of this
45 comments
Read more

Are product silos in a bank inevitable?

Silo thinking   is often frowned upon in the industry. It is often a synonym for bureaucratic processes and politics and in almost every article describing the threats of new innovative Fintech players on the banking industry, the strong bank product silos are put forward as one of the main blockages why incumbent banks are not able to (quickly) react to the changing customer expectations. Customers want solutions to their problems   and do not want to be bothered about the internal organisation of their bank. Most banks are however organized by product domain (daily banking, investments and lending) and by customer segmentation (retail banking, private banking, SMEs and corporates). This division is reflected both at business and IT side and almost automatically leads to the creation of silos. It is however difficult to reorganize a bank without creating new silos or introducing other types of issues and inefficiencies. An organization is never ideal and needs to take a number of cons
6 comments
Read more

RPA - The miracle solution for incumbent banks to bridge the automation gap with neo-banks?

Hypes and marketing buzz words are strongly present in the IT landscape. Often these are existing concepts, which have evolved technologically and are then renamed to a new term, as if it were a brand new technology or concept. If you want to understand and assess these new trends, it is important to   reduce the concepts to their essence and compare them with existing technologies , e.g. Integration (middleware) software   ensures that 2 separate applications or components can be integrated in an easy way. Of course, there is a huge evolution in the protocols, volumes of exchanged data, scalability, performance…​, but in essence the problem remains the same. Nonetheless, there have been multiple terms for integration software such as ETL, ESB, EAI, SOA, Service Mesh…​ Data storage software   ensures that data is stored in such a way that data is not lost and that there is some kind guaranteed consistency, maximum availability and scalability, easy retrieval and searching
26 comments
Read more

IoT - Revolution or Evolution in the Financial Services Industry

1. The IoT hype We have all heard about the   "Internet of Things" (IoT)   as this revolutionary new technology, which will radically change our lives. But is it really such a revolution and will it really have an impact on the Financial Services Industry? To refresh our memory, the Internet of Things (IoT) refers to any   object , which is able to   collect data and communicate and share this information (like condition, geolocation…​)   over the internet . This communication will often occur between 2 objects (i.e. not involving any human), which is often referred to as Machine-to-Machine (M2M) communication. Well known examples are home thermostats, home security systems, fitness and health monitors, wearables…​ This all seems futuristic, but   smartphones, tablets and smartwatches   can also be considered as IoT devices. More importantly, beside these futuristic visions of IoT, the smartphone will most likely continue to be the center of the connected devi
16 comments
Read more

Neobanks should find their niche to improve their profitability

The last 5 years dozens of so-called   neo- or challenger banks  (according to Exton Consulting 256 neobanks are in circulation today) have disrupted the banking landscape, by offering a fully digitized (cfr. "tech companies with a banking license"), very customer-centric, simple and fluent (e.g. possibility to become client and open an account in a few clicks) and low-cost product and service offering. While several of them are already valued at billions of euros (like Revolut, Monzo, Chime, N26, NuBank…​), very few of them are expected to be profitable in the coming years and even less are already profitable today (Accenture research shows that the average UK neobank loses $11 per user yearly). These challenger banks are typically confronted with increasing costs, while the margins generated per customer remain low (e.g. due to the offering of free products and services or above market-level saving account interest rates). While it’s obvious that disrupting the financial ma
7 comments
Read more

PFM, BFM, Financial Butler, Financial Cockpit, Account Aggregator…​ - Will the cumbersome administrative tasks on your financials finally be taken over by your financial institution?

1. Introduction Personal Financial Management   (PFM) refers to the software that helps users manage their money (budget, save and spend money). Therefore, it is often also called   Digital Money Management . In other words, PFM tools   help customers make sense of their money , i.e. they help customers follow, classify, remain informed and manage their Personal Finances. Personal Finance   used to be (or still is) a time-consuming effort , where people would manually input all their income and expenses in a self-developed spreadsheet, which would gradually be extended with additional calculations. Already for more than 20 years,   several software vendors aim to give a solution to this , by providing applications, websites and/or apps. These tools were never massively adopted, since they still required a lot of manual interventions (manual input of income and expense transaction, manual mapping transactions to categories…​) and lacked an integration in the day-to-da
5 comments
Read more

Can Augmented Reality make daily banking a more pleasant experience?

With the   increased competition in the financial services landscape (between banks/insurers, but also of new entrants like FinTechs and Telcos), customers are demanding and expecting a more innovative and fluent digital user experience. Unfortunately, most banks and insurers, with their product-oriented online and mobile platforms, are not known for their pleasant and fluent user experience. The   trend towards customer oriented services , like personal financial management (with functions like budget management, expense categorization, saving goals…​) and robo-advise, is already a big step in the right direction, but even then, managing financials is still considered to be a boring intangible and complex task for most people. Virtual (VR) and augmented reality (AR)   could bring a solution. These technologies provide a user experience which is   more intuitive, personalised and pleasant , as they introduce an element of   gamification   to the experience. Both VR and AR
9 comments
Read more

Beyond Imagination: The Rise and Evolution of Generative AI Tools

Generative AI   has revolutionized the way we create and interact with digital content. Since the launch of Dall-E in July 2022 and ChatGPT in November 2022, the field has seen unprecedented growth. This technology, initially popularized by OpenAI’s ChatGPT, has now been embraced by major tech players like Microsoft and Google, as well as a plethora of innovative startups. These advancements offer solutions for generating a diverse range of outputs including text, images, video, audio, and other media from simple prompts. The consumer now has a vast array of options based on their specific   output needs and use cases . From generic, large-scale, multi-modal models like OpenAI’s ChatGPT and Google’s Bard to specialized solutions tailored for specific use cases and sectors like finance and legal advice, the choices are vast and varied. For instance, in the financial sector, tools like BloombergGPT ( https://www.bloomberg.com/ ), FinGPT ( https://fin-gpt.org/ ), StockGPT ( https://www.as
2 comments
Read more

From app to super-app to personal assistant

In July of this year,   KBC bank   (the 2nd largest bank in Belgium) surprised many people, including many of us working in the banking industry, with their announcement that they bought the rights to   broadcast the highlights of soccer matches   in Belgium via their mobile app (a service called "Goal alert"). The days following this announcement the news was filled with experts, some of them categorizing it as a brilliant move, others claiming that KBC should better focus on its core mission. Independent of whether it is a good or bad strategic decision (the future will tell), it is clearly part of a much larger strategy of KBC to   convert their banking app into a super-app (all-in-one app) . Today you can already buy mobility tickets and cinema tickets and use other third-party services (like Monizze, eBox, PayPal…​) within the KBC app. Furthermore, end of last year, KBC announced opening up their app also to non-customers allowing them to also use these third-party servi
9 comments
Read more

Eco-systems - Welcome to a new cooperating world

Last week I attended the Digital Finance Summit conference in Brussels, organized by Fintech Belgium, B-Hive, Febelfin and EBF. A central theme of the summit was the cooperation between banks and Fintechs and more in general the rise of ecosystems. In the past I have written already about this topic in my blogs about "Transforming the bank to an Open API Ecosystem ( https://www.linkedin.com/pulse/transforming-bank-open-api-ecosystem-joris-lochy/ ) and "The war for direct customer contact - Banks should fight along!" ( https://www.linkedin.com/pulse/war-direct-customer-contact-banks-should-fight-along-joris-lochy/ ), but still I was surprised about the number of initiatives taken in this domain. In my last job at The Glue, I already had the pleasure to work on several interesting cases: TOCO   ( https://www.toco.eu ): bringing entrepreneurs, accountants and banks closer together, by supporting entrepreneurs and accountants in their daily admin (and in the f
9 comments
Read more