Microsoft Azure alternatives and competitors

I have a single 4GB model Linode and I use it as a personal server. I originally set it up to act as an email server, just for my own personal vanity domain. I don't use it for that anymore, but it's proved to be useful for many other things.

Right now, I run a Minecraft server on it and I also use it for a little bit of software development. I also use it as a jump host, if I need a stable place to SSH from my laptop to get to other online resources. It means that I only have a single point that I go through to get to the other stuff that I need.

I would say it's very important that Linode offers a relatively small, but well-focused set of cloud computing services because it differentiates them from AWS, Azure, and Google Cloud by being focused and by having a more personal touch. I could probably get the same compute power for a little bit less from other providers, but the value of the continuity and the high quality of support is worth it. It's rare that I need to reach out to support, but when I do, the support is great.

Even though I use it for personal use, some of the things that I use it for are in a software development capacity. For example, the ability to experiment with running my own Mercurial server, Git server, and source control servers on this machine have leveled up my own knowledge of those products in a way that I think a more managed solution wouldn't. Having the tools in-hand of just a Linux box that I can play with, and that I can wipe and reimage at will, is more useful than a physical machine by virtue of that virtual rewriting. It is also more useful than a more abstracted managed service, just in that I can get my hands dirty and do rapid experimentation.

Linode has helped me to accelerate innovation and even though I'm not using this in an enterprise way, it has improved the way I innovate with respect to personal stuff. For example, it has helped with the things that I'm trying to learn and the things that I'm trying to do. Setting up a Minecraft server is a good example of that. Being able to read some of the documentation that Linode has about setting up a Minecraft server on your VPS, and just learning and figuring stuff out, has been valuable.

The most valuable feature is the static IP address, which has been very helpful for being able to log into the same address over the course of more than a decade.

Another useful feature is being able to have multiple system images that I can play with. I mostly use Ubuntu Linux, but if I want to play with a new version of Ubuntu, I can pretty easily add that. It's been super useful to upgrade my system over the years.

I just recently started using the Object Storage and Backup features, as well. Having good backups for peace of mind and disaster recovery is very nice as well.

Object Storage has been key, for me. I don't have a strong notion of exactly when Linode introduced Object Storage, but it's been very useful for me, for instance, in backing up my Git server, in addition to the whole node backup. The fact that I can interact from the command line with the Linode Object service to back up specific datasets, is super cool. I know they didn't have it when I first started using Linode and I think it's been introduced fairly recently, within the last couple of years.

The Linode documentation is superb. 

I would like to see more seamless integration with backup, although it's pretty easy to do.

Having more on-demand features would be helpful. For example, if for a little while I wanted to have four Linodes instead of just a single one, it seems like it's a little bit more difficult than spinning up an EC2 instance in AWS. It isn't a lot harder, but it could be improved nonetheless.

I have been using Linode for more than a decade, since 2008 or 2009.

The stability is excellent. It's very rare that there is an interruption in the service. I think that the reboot notices that I get, that aren't related to me doing stuff, are less than one a year. I think, in some cases, it's been four or five years between needing to have any kind of stability-related events on the machine. I can't think of any outages in the entire course of me using it, that anything with any kind of significant impact.

Part of the thing that I value about the single Linode that I have is that it is a very steady, stable known quantity. I don't have to worry about all the institutional weight that I do with interacting with AWS, which I do a lot from work.

Thinking about it from an operations point of view, I think that I would have a bit of a harder time scaling in Linode than I would in AWS, but not a whole lot harder. Given the Kubernetes support, I would imagine that that makes the process even easier. That said, I have not tried Kubernetes so I really have insufficient data to be sure.

It's not likely that I'm going to significantly increase my usage in the future. I might bump up to a higher size if I find that I need more CPU or RAM. Or, I might play around with having two to four Linode nodes. But beyond that, it's unlikely that I'm going to expand much.

I will definitely keep using Linode for as long as it is as stable and reasonably priced as it is, but at a steady one machine for my personal purposes.

The support from Linode is great. All of the times that I've reached out, it's been through email or through the web portal. It's always felt good and it felt like the person responding understood what I was asking about and solved it very quickly.

In terms of flexibility and overall responsiveness, the support is very good to excellent. Certainly, everyone that I've interacted with, in the fairly rare occasions that I do need support, have been very knowledgeable about the product and very good at understanding what issues I'm having and how to solve them.

I would say compared to AWS, AWS support varies a lot in terms of responsiveness and whether you've got a paid support plan. Sometimes, it does take a fair bit of back-and-forth with AWS support to get to the crux of the problem. I've never felt that back-and-forth was as necessary, that we get to the crux of the problem and solve it much more quickly with Linode.

Prior to using Linode, all of the machines that I had managed were all physical. I had my own personal machines and machines that I built for work, but they were all physical PCs or other architectures that I had to actually open up a case, and if I needed more RAM, I had to put the sticks in myself.

In comparison to a physical server, Linode has definitely saved me money. I never want to build a server again. Basically, if you build a physical machine and it is obsolete within two to five years, you've got to buy and recreate the whole thing again. Generally, the hardware is going to get cheaper over time, but I think that unless I were really putting a microscope on getting the cheapest components for building, Linode will cost less.

In some contexts, albeit not mine with just one or a few machines, it would make sense to build them. However, not having to worry about it and just letting Linode take care of the hardware upgrades is probably saving me money. I don't know if it would save a very tightly tuned hardware IT team money, but that's a completely different scale than what I'm looking at.

Linode was my first experience with virtual cloud servers and virtual machines, in general. Not too long after I started using Linode, I did start doing more with VMware, with an on-premises, physical server hosting multiple virtual machines. It was not too long after that when I got into AWS for work.

I found the initial setup to be fairly straightforward. It's so long ago that the details are fuzzy but I recall that I set up the account, chose names for things, picked which size I wanted, and then launched it. Within, what at that time, was an astonishingly short amount of time, I was able to log into it. It's just gotten better from there.

I didn't require an implementation strategy, although I think that's peculiar to using it as a single thing for personal use. I had the notion that I wanted a persistent Linux machine that was always on, and that I could get to from anywhere, and Linode seemed to fit the bill.

Over the years, I've used it for a lot of different purposes and it's adapted well to that. So I would say in this case, I didn't need a whole lot of planning. If I were to use Linode for a more complex deployment, I would want to plan it out, figure out what the costs are, figure out the network topology, and the other relevant details.

Linode offers worldwide coverage via multiple data centers, although I don't personally need that. It's a very attractive feature for sure, but since I just have the one virtual private server and it's just for me and for my friends connecting to it, I want it to be geographically close to me to have a low ping. I think it's located in New Jersey, and that's good for me, being in New England. While global coverage is not super important to me, in so far as it contributes to the health of Linode in general, I'm all for it.

The pricing model is simple, and that's one of the reasons that I've stuck with Linode for so long. When I was on the $20 Linode, I knew for a fact that on the first of every month, my credit card would get charged $20. That meant my Linode was available constantly.

At this point, I'm paying more like $35 per month for a slightly bigger machine, and the backup, and the object service. But again, I know that it's exactly $35 every month and I can budget for it. The simplicity and the consistency of that billing and pricing are quite valuable to me. Whereas with AWS, it's a crapshoot. The on-demand pricing means it's flexible and I only pay for what I use, but it's also much less predictable.

It is tough to determine whether using Linode has saved me money compared to what I would pay with other cloud providers. I don't think it has on a pure numbers basis, but in opportunity cost and higher-level budget planning, I think that the consistency probably has saved me money. I would have spent more time trying things, allocating things that I might not need, and so on. Ultimately, it saved me capital in the long run but it is not necessarily something that I can put a dollar figure on.

In comparison to everything else, predictability is the key aspect of the pricing model. With it being a known quantity that I can budget for every month, it frees up brain cycles to do everything else.

In my personal capacity, at this point, I have my Linode and I have a personal AWS account, and I use them for different purposes, but to similar degrees, or similar magnitudes. I find that just looking at it from a strict CPU and RAM horsepower perspective, EC2 is just marginally cheaper, but there are different features that I value in different places.

I took a brief look at some other things like Azure, Google Cloud Engine, and DigitalOcean, and I found that when I was looking, and this was probably about five or six years ago, that a lot of things that I wanted were pretty comparable in terms of capabilities and pricing. A lot of it came down to what I valued, in terms of the positioning, and support, and documentation, where I very much like Linode's documentation, especially, and support.

Of the others that I evaluated, DigitalOcean seemed the friendliest. And then, AWS and Azure were the behemoths, the 800-pound gorillas in the room.

The biggest lesson that I have learned from using Linode is the oldest lesson, which is just that a virtual cloud server has the availability and the flexibility that I couldn't get from physical at the time, or even now, for that matter. It's a key component in having something that's useful, having a machine that you can log into and do things on, in a consistent way, regardless of where I am or even what machine I'm connecting to it from.

My key advice for anybody who is looking into Linode would be just to dive right in. Pick it up and play around with it and if you find that it's not for you, try something else. But if you find that you love it, keep going.

In summary, Linode is a good product and I've been extremely satisfied with it for exactly the purposes I use it for. I have been pleased with it since I started using it.

I would rate this solution a ten out of ten.

We are using GoogleCloud for hosting a SaaS platform.  

The most valuable feature of the product is that it has a very powerful command line.  

It is difficult to say what should be improved because, obviously, they have made some major improvements in the organization of how you do things — such as the way you set up a server. Google has made a lot of effort to try and catch up to the competition in the area of ease-of-use because that would have been my one complaint: that you have got to be quite technical to understand some of the ways that things are done. Azure and Heroku are number one in ease-of-use and they make it very easy. Google has done a lot of work to alleviate that objection and to catch up with Heroku and Azure. But the people that have the most to say about the ease-of-use would be the guys using it. For the developers we have, they like the power and the control that Google gives them.  

So I can not actually answer what exactly has to improve for developers to be more satisfied at this point because they seem quite satisfied with it already. I do not get any complaints from the guys. They are the ones using it every day and I do not use it on a daily basis so I really can not comment in that sense.  

I could say that it can be easier to use for people that do not have the same level of technical skills, but even that has improved a lot with their upgrade to the user documentation.  

I have experience with Google Cloud for about two years now.  

It is extremely stable. I have no negative feedback and no complaints.  

Google Cloud is very scalable. The use of Docker and Kubernetes has really made it extremely scalable. Google's implementation of Kubernetes is excellent.  

At this stage mainly we have developers and dev-ops using the product and it is a team of about 25 guys. We could expand that at any time.  

The initial setup is somewhat more complicated than competing products. I was not directly involved in that capacity so I can not provide details. But also complexity sometimes leads to opportunity as far as customizing performance. The people who are working with the product directly like the ability to fine-tune more than they want simplicity.  

Initially, I was just doing research. I was keeping an open mind and looking at all of the possibilities. I do not think it is quite right to call it "evaluate" when you just do a feasibility study. We did take a closer look at a few solutions like Azure, Heroku, and Google Cloud. We dabbled a little bit with some of them. We eliminated Azure because we were concerned about the support for Mongo DB which is part of our technology stack. Obviously Azure has changed a lot since.  

We trimmed that group to Heroku and Google Cloud which are technically both platforms of high-quality. Heroku is easier to deal with. Using Google Cloud, you have to build a bit of experience with the product because it is not easy if you do not understand how to do things. Heroku makes it a lot easier for you.  

The reason we went with Google Cloud had to do with two things. Number one was cost, and number two is that Google supported everything we use. We had to control the costs initially, so Heroku was pretty much out the door almost immediately. It was a competitive product but it was too expensive. An end-user would not know where a platform is hosted, and they would not care. For an end-user, they go on your website or on your SaaS platform and there is no difference in the experience whether you are hosted with Google or Heroku or AWS. It makes no sense for them to worry about that. But the cost ends up being an important component of the decision for the service company.  

I think the point is that it is very difficult for companies these days to decide between Heroku, Azure and Google Cloud. They all have data centers in the right places in the UK. There is very little that differentiates any of them. Heroku obviously stands out because they are a very stable platform and they do all the hard work for you. If you do not have the expertise to go with a less expensive more labor-intensive solution, then you would go with Heroku and pay more.  

We have a development center in Manila with very experienced guys and they love Google Cloud. It gives them everything they need and everything that is required for a big, fast platform — like the ability to use clustering. I think all of the solutions support Java and ATC (Advanced Analytics Technology). But we have not had any issues since we started on Google Cloud, so we are happy with the direction we have taken.  

The lesson I learned from adopting Google Cloud is that you should do more training before you commit to it.  

On a scale from one to ten where one is the worst and ten is the best, I would rate Google Cloud as an eight-out-of-ten. You can not have a perfect platform. There is always room for improvement and something to add.  

That rating is really because of the feedback I get from the team. I get good feedback from the guys. But it is not really fair of me to give any product a ten if I am not more intimate with it in daily use. I could just as well give it a one if I were totally ignorant of the product, but that would not mean anything.  

While I cannot say for certain, I believe that we are using the latest version. 

We primarily use the solution to rent servers for storing certain commercial applications.

I especially like the flexibility and scalability of the solution. It is totally scalable. 

While feasible, custom configuration will be more time consuming than standard, although we have not encountered many instances which required us to seek support or advice.

I believe we have been using Amazon AWS for more than 10 years.

The solution is absolutely stable. This is one of its best features. 

The solution is absolutely scalable. 

Amazon allows us to scale up and then down, something important to one of our customers who was in need of temporary increases in the throughput provided to the servers. This allowed us to meet the client's needs for the days or weeks that they required more dynamically located servers, after which we were able to scale down. This we were able to do through Amazon. This was difficult to accomplish beforehand, as the client had private servers for which he was forced to buy machines which he would subsequently keep. 

I cannot comment on Amazon's technical support, as we have not made use of it. 

We did use other solutions prior to Amazon AWS. We made use of local service and dealt with projects involving Google and Microsoft. We also used Microsoft Azure. 

Not long ago  we used Microsoft Azure, though this is necessary with some of our projects. We have different projects which vary with the customer's specifications. Some utilize Azure, although most require the use of Amazon. 

When comparing Microsoft Azure with Amazon AWS, I do not see much disparity. It really comes down to a business choice. If the customer is familiar with Microsoft, then the testing team maintaining the product will need to be acquainted with it as well and its ongoing use is required. Similarly, Amazon will continue to be employed if this is already the case. As such, the difference betwen the solutions does not come down to considerations of a technical nature as they are largely similar. The primary consideration is one of business, the use of one solution and provider over another. 

When it comes to standard configuration, the installation is quick, usually taking one or two days to complete. Custom configuration, while feasible, takes somewhat longer. So far, we have not had many instances in which we required support or advice concerning custom configurations. 

The technical team would be in a better position than I to address any technical issues involved in the setup. From my perspective as a project manager, I feel what we have to be sufficiently good. There is much advertising, information on the advantages of the product and guides available. 

Installation was carried out by our own internal integration team, not externally outsourced. I did not handle it myself. It was done by a team specialist. 

The technical team responsible for the deployment consists primarily of engineers. 

I cannot comment on whether we have seen an ROI, return on investment, as I do not possess this information.

The licensing cost varies with the project involved. Certain projects run around $6,000 per month, some less and others more. We handled many projects, each with its own complexities and specifications. The price ranges of the licenses varies with the complexity of the project. 

Broadly speaking, there is a need to rely on specialists for properly setting up one's accounts and addressing his needs. This is not specific to Amazon, however, but is something prevalent with all providers.

I have assumed the role of both customer and integrator. In the past, I worked as a project manager with different projects employing Amazon products, services and software. 

For the most part, the solutions I used have been public, not private, such as AWS cloud. 

The number of users of the solution varies with the individual project. This can range from 20 to 200 to 500 users. 

Our teams have undertaken every role, be them architecture, development, design or testing. They are all internally integrated. 

I am a fan of Amazon products and generally recommend them to others. Of course, we employ Azure and Google products when the customer specifically requests these. 

Since all products have room for improvement, even when this is not apparent to me, I rate Amazon AWS as a nine out of ten.

We are a managed service provider and IBM Cloud is one of the products that we use to provide services for our clients. We have a large number of customers who are in different industries and some of them are moving from a traditional data center to the cloud.

There are several advantages to migrating to a cloud-based data center. Some of these are that you don't need 24/7 power or power backup units, local staff for maintenance, or to handle your own security.

There is also the important point that servers are available on-demand. For example, I may need 100 servers for a client for a short time, but after which they are no longer needed. This ability to scale up or scale down using the cloud is helpful. By contrast, in an on-premises solution, if you need more power then you have to purchase it.

One of the features that I really like about IBM Cloud is the flexibility where you can order your own hardware. Customers do not want to have a full data center but they can order virtual machines, use solutions like VMware vCenter and NSX, and they can design their own network. 

This solution is user-friendly and easy to manage.

There is not a lot of support for this solution, which is something that needs to be improved. 

We have been using IBM Cloud for a few years.

IBM Cloud is stable and our customers haven't had any issues.

This solution is very scalable. IBM took over a company called SoftLayer, which has resources across the globe. They are available in the UK, US, Australia, India, Asia, and everywhere.

There is not a lot of support for IBM Cloud. The response from them and the ETA for resolving issues is too slow, and also, the SLA is too expensive. If you want to have premium customer support then you have to pay more.

Obviously, they will still make the best effort according to the SLA. Whatever engineer becomes available will give good support but it is a very slow process.

For example, something that is a priority-one issue for you will not necessarily be of the same urgency for their support. Rather, they will go as per the SLA, which is based on how much you pay for support. If you are a gold partner then you can just pick up the phone, tell them the problem, an engineer will be assigned and they will just start troubleshooting. For most organizations, it doesn't work this way.

I have worked with AWS and Azure clouds. IBM Cloud is the cheapest of these three, and I found it to be more flexible in terms of third-party integration.

The initial setup is straightforward and not very complex. The fact that you can make your own infrastructure is the best part.

Being a network architect, I set up all of these things on my own. There are different vendors who can support you, based on your requirements.

The price of IBM Cloud is very cheap compared to competitors AWS and Azure.

The licensing fees are billed on a monthly basis and some costs vary between cities. For example, data centers in smaller cities will cost less. The savings in operating costs of a remote data center are passed on to the customer.

Premium support is available for an additional cost.

The suitability of this solution depends on your organizational requirements and how you want to function, as well as the policies. One example is that a lot of financial institutions do not want to migrate to the cloud for security reasons. Some organizations prefer one provider over another, such as AWS. There is no hard and fast rule that says you have to use only the IBM Cloud. The option is up to the customer.

Overall, I would stay that IBM Cloud is stable, straightforward, and is a good service. I recommend it. My main complaint is about support, which is tied to pricing.

I would rate this solution an eight out of ten.

Right now we're using two cloud platforms, one being Oracle Cloud Platform and the other Microsoft Azure. We've been using Oracle Cloud Platform since 2018 and Azure since 2015.

We use Oracle Cloud Platform with the OCI Infrastructure as a Service offerings, and we're essentially using it for the compute instances and nothing else. We have deployed some servers running Oracle Linux which are being used as our main development environment. Then, we also have servers for our UAT (User Acceptance Testing) environment. Neither of these use cases are for production environments, and they are only used for development and testing.

To access Oracle Cloud Platform, we have a VPN site-to-site connection from our site to Oracle Cloud. We don't have the FastConnect option because we don't have a very good reason to use it.

As for users, we typically have around 20 people who are connected to the Oracle environment and who do their work on it. Regarding maintenance, we have three people on technical staff for that, who are DBAs and architects.

It hasn't improved our organization in a direct, or highly visible, way because we only use it for development and testing. However, I enjoy the ease with which we are able to deploy Oracle Linux server instances and other Oracle products. 

I like having the ability to easily run Oracle Linux server instances and to deploy Oracle Middleware and WebLogic servers. Oracle's Infrastructure as a Service products are also very useful, and we're using those right now within Oracle OCI.

I think that there could be a more user-friendly environment when it comes to the options that Oracle presents through the Oracle Cloud Platform.

Also, when you compare it to Microsoft Azure Marketplace, there aren't as many options to choose from, and some features are not available in some operating systems. In particular, I'd like to see more readily-available features such as Fusion Middleware as a service, in the same way that they have WebLogic as a service within OCI.

I have been using Oracle Cloud Platform for the last three years.

It's very stable, especially when it comes to accessing the Oracle Cloud Platform through VPN.

It's very scalable and we haven't had any issues in that regard.

We have not previously used a similar solution to Oracle Cloud Platform. We started with Oracle Cloud Platform because we were already using Oracle products and we decided that deploying Oracle products on Oracle Cloud Platform would be a more sensible idea than deploying on other cloud solutions.

The setup was very straightforward and it only took one to two days for the initial deployment.

We used Oracle consultants through Oracle's ACS, along with three technical personnel including DBAs and architects.

On a monthly basis, we pay around $2,500 for licensing fees, with no additional costs. We do pay extra licensing for other on-premises solutions, but these are not related to Oracle Cloud Platform.

We're not using that many products within Oracle Cloud Platform, but nevertheless I can recommend it. Especially when it comes to the Database as a Service, which we use a lot. I can easily vouch for that if you are looking for good database performance.

In addition, if any companies or clients are currently using Oracle Database on-premises, I can certainly recommend that they migrate to Oracle Cloud Database. 

I would rate Oracle Cloud Platform an eight out of ten.