What is our primary use case?
AWS Lambda has serverless programming, like Logic Apps from Azure. You just configure the run-time and then they start coding. It is event-driven. It started with my obtaining Salesforce. Salesforce is a low-code and non-code program and totally SAS. Everything starts from the event, from the trigger. You get the trigger and you work at the program. You have some other models, maybe faster or fancier models. But in my opinion, this kind of program is started by locating the system and identifying where the trigger and entry point of the program are. Then you get the full advantage of the program. You don't need to worry about any infrastructure.
I think this is the future. Compared with the EC2, you don't have to pay anything if you don't run it. Otherwise, with EC2 when our client provisions the system and the instances, you always have to pay. There are other tremendous advantages, like flexibility. After you provision EC2 you can write something that does not totally follow the cloud convention. You use it to provision the container. With the program you need to have those 10 principles of cloud computing. Especially recently, within the past four or five years, I have gotten away from DevOps, or the software development life cycle. Even though I researched the product portfolio from DevOps and then the life cycle for DevOps, I try to position myself as an architect with hands-on experience.
In my opinion, Lambda is very similar to Salesforce, which is the original for the SaaS platform and is an extremely low-code environment. With Microsoft and AWS you can say, "Okay. You can choose whatever language you need to make it even more flexible."
Everything is the cloud. Lambda is a fully managed service. If you want to do it either as a private cloud or on-premise, I'm sure you can do that, too. But I don't know how to manage the pricing structure. But then you've lost the point of Lambda because if you do not use it, you do not pay. Again, I just want to emphasize, I'm not a Lambda expert. But, logically thinking, the big advantage of serverless programming for the customer is that you just use it and pay. Pay and go. You don't need to provision anything.
All my experience with AWS Azure is on the public cloud. We do not get too deep. In IBM we do. When we do sales training we always get the private cloud on-premise. There are many reasons for this. One reason is that IBM lost the battle for the public cloud so we get into it much deeper. We go to the enterprise and we can deploy programs to your data center and offices. But for the tech data for AWS and Azure, we are all using the public cloud as a showcase when we talk to the customer and to the retailer.
What is most valuable?
The number one feature with AWS Lambda is that it is fully managed. From the developer's perspective, you get the coding much more easily. Now many situations are not using code. You plug in, assemble it, and configure it. Lambda makes it low-code. I come from being a Java certified developer for 15 years. You configure the environment for deployment just like in DevOps. That was always the most challenging part as a developer. You identified when to trigger it. If the program can't facilitate it, then 80% is gone. With 20% you just Lose Syntax. You can use Lose Syntax with any programming language as a reference finding out the variables, the statements, the loop, and what other kinds of things you can do. Just follow that to where you can plot it into your business system.
They might think to have the business benefits say, "Hey, if you don't like it, no need to pay." So, potentially, you can save. If the future is going to be serverless, that's what I think the future of something like Salesforce will be. Programming is getting much easier and does not need a lot of configuration because step-by-step abstraction starts from the infrastructure service. You can replace your hardware, but you still need to do a lot of things in the abstract. The environment now is totally fully managed. I'm not sure if we're totally aligned there. I always talk against those aspects in the Salesforce situation. But I believe Lambda is a comparable peer, apples to apples.
What needs improvement?
I can only speak from the user experience. Salesforce integrates SharesPost efficiently. How? They say, "Okay, I invented another language called Apex. Forget about anything else. This is my language." The benefit of this language is that everything is simplified. Your system is super easy to maintain. But AWS then assures you that they are flexible, that they have a collection of 10 or 20 languages, and you just choose your environment and range. That's the reason I appreciate Salesforce. They always make things easier. They have their loop reasoning because they are a different kind of company. Microsoft and AWS really get the full spec. They want to own the business. But Salesforce data wants the simplest way.
So, this is my understanding and unique experience.
I think that perhaps Lambda could explore its functionality more.
For how long have I used the solution?
I have been using AWS Lambda for a few months.
What do I think about the stability of the solution?
I didn't explore enough information to evaluate that.
What do I think about the scalability of the solution?
I didn't experience the scalability personally, only from my reading. Amazon takes care of the scalability. That's the right way. It's automatic and it's fully managed. That's one benefit of Lambda.
We have all kinds and sizes of resellers. There are large enterprises and small businesses. It's different. And some of them are product based, they are creating their own products. Some of them are consultant based. It's really different. Tech data is different vs. a business model.
How are customer service and technical support?
I contacted support many times. My experience was very little and I just saw how Lambda was working, to try to understand if it is okay.
Which solution did I use previously and why did I switch?
I don't actually use AWS Lambda. I'm a distributor. I try to explain solutions to the vendor. I previously used Salesforce Apex. I use the Azure Logic App service.
Salesforce does not have so many options to choose from, such as Java or C++. Salesforce said, lets invent a language. They call it "invent" but actually they just made a simplified edition of Java and eliminated a lot of complex features. Now all the syntax is the same. Salesforce is a business company. They focus on business solutions development and they make the customer's lifecycle development simple. AWS really does not stick to any business because they are a technology company.
Let me explain the similar things that Lambda has to Salesforce. When you get the event you have to see our form. With the sales approval process, if you have the 50% to get to the half million and above, you need the vice president to get the approval. You can use this trigger based serverless program. All you want to do is to write down the logic and then put it under the trigger of whenever a certain number changes in the half billion, and then you need to do the multiple steps.
This kind of programming is easily defined in the business. All you need to do is get the logic done, get it tested, see the steps you are doing, and then fix up the errors. As for Lambda, as I said, I've just experienced two very simple examples in the AWS, but they were the same thing.
Logic App and Lambda should be doing the same thing - fully managed coding. You focus on the logic triggered by the certain events. And there are other additions within the Lambda family. It can be scheduled as a batch job. I don't think it's originally lack of motivation from the serverless. The serverless is from the trigger.
How was the initial setup?
The initial setup is straightforward. If you follow a 30 or 45 minute lab, it seems pretty clear.
What other advice do I have?
Everybody should check out AWS Lambda. That's why I didn't explore much and it was at the top of my list. This is a fully managed model. The number one. This is for the future. In the future, many of the EC2 applications may be replaced by Lambda. If I started something from scratch, I would try to use Lambda. It's much simpler. It can simplify a lot. If you add the scalability into the picture, it could have 80% or 90% of the complexity. They are very important. All the servlets are very important from a cloud computing perspective.
On a scale of one to ten, I would rate AWS Lambda an eight.
I am a fan of the no-code, low-code if you consistently improve to make it even simpler. Maybe they could do something to simplify the language. I'm not sure if Lambda has the code for the Microsoft Logic App, which means they can eliminate most of the code and everything becomes drag and a drop. Because they eliminated those "if errors." They have those kinds of functions. I think mostly because I have not explored the whole portfolio of AWS. I believe there is a full suite of them.
I believe their full suite of the service is complemented with Lambda. But I do believe the competition is going to make it simple with low-code, no-code. There is no-code, low-code and also no infrastructure. That is going to be the key. Also, maybe you can have the Lambda ecosystem and have some component of the module built above the Lambda so that people can make graphing and plotting even easier. This is not just any software, you get the module there which is much better. But AWS is big enough to neutralize the ecosystem. I believe it will come but the people don't have the patience to start from scratch these days.
Which deployment model are you using for this solution?