Share your experience using Amazon Lightsail

We use Amazon Lightsail for hosting security components in-house and AI modules.

When you need to add some custom rules to Amazon Lightsail, the auto scaling feature is a bit difficult. You can create custom rules while using a Lambda function or EC2 EKS cluster, but it's not feasible to do so in Amazon Lightsail.

We are a service-based organization, and we have been using Amazon Lightsail for more than three years.

If your organization's DevOps team has good troubleshooting skills, you can easily manage stability issues by creating custom scripts.

I rate the solution nine and a half out of ten for stability.

You cannot implement a Lambda function and custom rules in Amazon Lightsail. When an application's specific process or service takes time and too much load, taking a network-based or IOPS-based autoscaling is not feasible.

I rate the solution seven and a half out of ten for scalability.

The solution's technical support team assists you as per the SLAs. The support gets a bit delayed for beginners.

I rate the solution’s technical support six and a half out of ten.

Neutral

The solution's initial setup is easy and takes around one to two days for someone with a background in SDA or Linux administration. Since it is UI-based, the solution's installation is easy to learn.

The solution's deployment takes five to ten minutes. We implement zero-downtime deployment as well.

Amazon Lightsail is a bit cheaper than Amazon EC2. Amazon Lightsail is a cheap solution for startups.

Amazon Lightsail makes it easy to deploy any application on any cloud. You do not need to set up the whole website. Instead, you can just drag and drop your applications. I would recommend Amazon Lightsail to users who do not have much traffic. Amazon Lightsail is a bit cheaper than Amazon EC2.

Overall, I rate the solution an eight out of ten.

The use case mainly was to deploy a database that there was no option for the platform as a service. It is a database based on a graph, the name is Neo4j. So in AWS, I didn't find the RDS of this database. So I had to create a virtual machine in Lightsail and deploy the Neo4j database on it. 

I did it for the test and production environment, and then it allowed me to create firewall rules and a lot of management networks, but they are pretty simplified. It's not like EC2 where you can have all of those virtual interfaces and virtual networks. It's more like an instance that is standalone. But at least for these small purposes, it attended very well. 

Lightsail has simplified the process of setting up and managing a virtual private server a lot. 

Because I'm able to do it in a few clicks and have the data machine deployed. Like ten clicks is very different, very simplified, then EC2, in that you have to configure a lot of virtual networks, disk management, and a lot of other points regarding the virtual machine environment. 

And in Lightsail, they simplified to do it in as few clicks as possible. So, around ten clicks, you have a virtual machine deployed.

I also leverage it for basic monitoring capabilities. I've set up some basic alerts, for example, to email me if the operating system shuts down, to be aware. But that's the extent of my monitoring features.

There were templates for Ubuntu, for instance, that I used. I used it if I remember, going to 20.04, I guess. It was ready, and I just used the template and it worked well. There is a template for Windows Server, but I didn't test them. 

Also, the configurations over the firewall in the console work very good for the purpose that we need to isolate it from the internet but allow the connection from specific IP addresses in specific ports. 

So the filtering in the Lightsail Firewall works good. And I am able to isolate the virtual machine well from the internet in order to avoid DDoS attacks and stuff like that.

To improve, I feel the disk management is too basic. I can't easily partition the disk for the operating system. I have to follow the main configuration that is already set up in the template. 

To partition the disk, I would need to deploy the operating system first and then make changes, but it's not an ideal method. I would recommend improving the disk management.

I have been using it for the last one and a half years, I deployed some virtual machines in Linux, and it was pretty easy to use.

It is very stable, I would rate the stability a ten out of ten. I had no issue regarding stability at any time.

I haven't scaled yet, but I'm aware that there is the option to change the size of your virtual machine to a larger one. 

Scaling would require you to shut down the operating system, and then you can change the size with just a few clicks. After restarting, the machine is scaled to a higher level of sizing.

However, the scalability is not too good. I would rate it a seven out of ten because there are not too many options for sizing. So if we need to scale, we will have to move to a much bigger size than we are, and perhaps that will become more expensive than if we had more sizing options for our virtual machine.

As long as it's just for a database, there are not too many users, maybe only those two, me and the application manager.

Lightsail also offers options for Virtual Desktop Infrastructure (VDI), so it could be used to connect people from an organization to work in VDI sessions. 

As far as I remember, Lightsail doesn't provide support. There wasn't any at the time I deployed. If there is now, I have never used the support.

I'm not a hard user in cloud, but I've used Lambda, API Gateway, AppRunner, and also Lightsail for deploying virtual machines, all in AWS. 

In Azure, mainly RDS databases like SQL Server, and the common virtual machine is for infrastructure as a service. 

For the platform as a service, we are just starting to have evaluations, and most of them were in AWS and involved AppRunner.

AppRunner from AWS is the service that I can use to run, for instance, an HTTP website. It will point the website in GitHub, for instance, to AppRunner, and it will deploy the website. It has embedded the runtime and all the libraries that are needed to deploy the website.

For web hosting, I have used AppRunner because I was trying to go in a serverless application, and AppRunner, at least for the application, was able to do everything in a serverless way. I only use it for the database because I had no option to select that database in an RDS instance. So I had to create the virtual machine and deploy the Neo4j database on them. But for application and web hosting purposes, I didn't use Lightsail.

The initial setup is very simple and straightforward.

Overall, I would rate my experience with the initial setup a nine out of ten, with ten being easy to set up. 

At this moment, we are still mainly on-premises. We have a proprietary data center, but we are gradually moving to the cloud to implement new and modern applications in the cloud. This is a long-term journey because we are in the industrial sector, which does not require scalability or elasticity.

So, it's common that we can support everything from our premises. However, to leverage some benefits from the cloud, for instance, this website that I'm probably sharing in AppRunner, is mainly for customers, not for internal purposes. When talking about external people accessing our website, it is much better to have it in the cloud than on our own premises.

I had help from a consultancy partner who was working with me on the project, but it was mainly dedicated to the database configuration itself because it's a very specific database. My experience with that was not too much, so it helped me to configure the Neo4j database.

It was a two-person team. It was me and one more person. And he was a database administrator. But he was also the architect of the solution itself, the application. So he was aware of how to configure this database to meet the demands of the application.

The deployment was two days, not more than that.

The deployment process was using the templates from Lightsail. And after that, I went to the Neo4j Installer and did the configuration for this database. It was very easy. To deploy was two days because one day to configure everything regarding the operating system and another day to configure the database.

One person is enough for the maintenance with Lightsail. That's enough because Lightsail itself is not a big deal in terms of management. It can provide everything you need in a few clicks and make everything easy.

The prices are better than EC2 machines, for instance, because they are simplified management, but it works well, and I am able to run everything that I need on it.

The flat-rate pricing model of Lightsail allows you to select the size of the virtual machine, and although there aren't too many options, at the time I chose there were about five sizes available. I selected a middle-size option, which costs around $20 a month. This is a very good price, especially for the US East region. 

For my budget, this was a suitable option because, in comparison, EC2 from AWS would have been much more expensive. 

Although EC2 offers a lot of additional benefits in terms of configuration, if I don't require those extra configurations and features for our environment, and I just need to deploy it with some protection from the internet, Lightsail works well.

I opted for Lightsail in comparison to EC2. 

It's mainly the management console. In Lightsail, it's very simplified. I have a lot of options to configure our virtual machine infrastructure as a service in EC2. 

Lightsail is more suitable if you have a fast need to have a virtual machine online for any simple purpose, making it a good starting point for users not familiar with configuring a virtual machine environment. 

Lightsail offers a good starting point for people. 

The price is the main attractive point. It is, in my opinion, much better than EC2 for virtual machines. So I, for sure, would recommend it.

Overall, I would rate the solution a nine out of ten. It's a very good solution.