While integrating Google App Engine with other services, I don't recall any specific challenges. The main drawback with Google App Engine's standard environment was its restrictions. We could not work with file systems, run shell scripts from the environment, or use WebSockets. These issues could be resolved using the Flex environment, but that increased the cost. The Flex environment requires one service instance to always be running, which further increases the service cost. It would be beneficial if Flex services could scale down to zero instances during periods of no traffic. This would solve many issues where we had to make compromises because we didn't want to maintain a constantly running instance while still needing the flexibility of WebSockets or low-level system APIs that weren't available in the standard environment. Though I have experience with AWS, it's not comparable as I only used EC2 services, where you manage your own VM independently.
There are two versions of Google App Engine: flexible and standard versions. I think they can improve by having only one version. The advanced version has better features, better scalability, and better performance, but it costs more. Instead of having both versions, they should have one system with standard pricing. I would prefer to have only the advanced version from Google App Engine instead of two systems.
The areas of Google App Engine that I would to improve or enhance include its allowance for complete end-to-end deployment and scalability; however, it is manageable only for a few languages. For instance, it doesn't support languages C and C++, only basic support for Node.js, Java, and Ruby. It's not a comprehensive solution for all scenarios. Moreover, the security feature is based on IAM roles, but it should ideally be based on Active Directory (AD) roles. For IAM-based roles, we need to add the proper users and provide all security permissions manually. In an AD-based model, we would simply add users to a specific group, and all permissions would be inherited.
The main areas for improvement in Google App Engine include security restrictions and deployment processes. The whitelist and blacklist of APIs can be a deal-breaker due to security concerns, and the deployment process is chaotic. A steep learning curve also exists due to the non-standard Google way of doing things, which could deter people from adapting this solution. Customization needs improvement as the current limitation to Google's way restricts flexibility.
There is limited customization because the sandbox environment restricts it. There are issues with cold starting; apps may take time to wake up if idle for a few minutes. The pricing is higher compared to alternatives like Kubernetes ( /products/kubernetes-reviews ) and Compute Engine.
The support for the Indian region is not as good as compared to the support that is offered to the regions in Europe. My company faces a lot of problems with the support team, especially when Google's technical team is not able to help us immediately when we are stuck with something related to the tool. The support of the product is an area where improvements are required. Google App Engine is a good product, and its support needs to be improved as it is a new technology that few people know about. There is a need to spread some awareness about the product while also ensuring that some online forums related to the product need to be updated with a lot of documentation related to the tool.
Pre-Sales Solution Architect Manager at Qi Network
Reseller
Top 10
2023-07-18T12:45:07Z
Jul 18, 2023
Some features of runtime don't work well in App Engine. Sometimes we must tune it ourselves or ask Google’s team to improve it. It is the most important thing. Our clients always complain about it.
The documentation is an area that could definitely be improved. Another area of improvement would be the community which is poor and Google could do a lot better on that front. I know that my team members who have been working with the solution for just a year or two find the architecture to be quite difficult.
Programma en project manager at a consultancy with 51-200 employees
Real User
2022-04-07T05:11:18Z
Apr 7, 2022
I would like a simpler deployment tool on laptops. It is a bit complicated at the moment. We know how to do it, but it could be easier to deploy it on laptops.
DirectorManaging Partner at a tech services company with 11-50 employees
Real User
2021-11-07T09:41:39Z
Nov 7, 2021
The only concern is that there is a number of the offerings which are built on their own proprietary technologies. With some of the offerings in Google Cloud, it's difficult to have a path to migrate to other cloud providers.
Teaching Assistant at a government with 201-500 employees
Real User
2020-04-23T10:13:34Z
Apr 23, 2020
Everything so far has gone smoothly, I've never had a struggle with the solution. In terms of product improvement, when it comes to billing, I think they should include something that would give the client or user an indication of what's happening so they can be aware of how pricing is being managed.
It would be helpful if they could improve the user interface. It could use some modifications. There needs to be more directions in terms of how to use the solution.
Google App Engine is a Platform-as-a-Service (Paas) provider that equips web application developers with all the resources and tools that they need to develop, test, and run their applications on Google's infrastructure. Everything is built into the kit, so with one download of the SDK, you'll be well on your way to first-rate apps.
While integrating Google App Engine with other services, I don't recall any specific challenges. The main drawback with Google App Engine's standard environment was its restrictions. We could not work with file systems, run shell scripts from the environment, or use WebSockets. These issues could be resolved using the Flex environment, but that increased the cost. The Flex environment requires one service instance to always be running, which further increases the service cost. It would be beneficial if Flex services could scale down to zero instances during periods of no traffic. This would solve many issues where we had to make compromises because we didn't want to maintain a constantly running instance while still needing the flexibility of WebSockets or low-level system APIs that weren't available in the standard environment. Though I have experience with AWS, it's not comparable as I only used EC2 services, where you manage your own VM independently.
There are two versions of Google App Engine: flexible and standard versions. I think they can improve by having only one version. The advanced version has better features, better scalability, and better performance, but it costs more. Instead of having both versions, they should have one system with standard pricing. I would prefer to have only the advanced version from Google App Engine instead of two systems.
The areas of Google App Engine that I would to improve or enhance include its allowance for complete end-to-end deployment and scalability; however, it is manageable only for a few languages. For instance, it doesn't support languages C and C++, only basic support for Node.js, Java, and Ruby. It's not a comprehensive solution for all scenarios. Moreover, the security feature is based on IAM roles, but it should ideally be based on Active Directory (AD) roles. For IAM-based roles, we need to add the proper users and provide all security permissions manually. In an AD-based model, we would simply add users to a specific group, and all permissions would be inherited.
The main areas for improvement in Google App Engine include security restrictions and deployment processes. The whitelist and blacklist of APIs can be a deal-breaker due to security concerns, and the deployment process is chaotic. A steep learning curve also exists due to the non-standard Google way of doing things, which could deter people from adapting this solution. Customization needs improvement as the current limitation to Google's way restricts flexibility.
There is limited customization because the sandbox environment restricts it. There are issues with cold starting; apps may take time to wake up if idle for a few minutes. The pricing is higher compared to alternatives like Kubernetes ( /products/kubernetes-reviews ) and Compute Engine.
The product's price is an area of concern where improvements are required.
The support for the Indian region is not as good as compared to the support that is offered to the regions in Europe. My company faces a lot of problems with the support team, especially when Google's technical team is not able to help us immediately when we are stuck with something related to the tool. The support of the product is an area where improvements are required. Google App Engine is a good product, and its support needs to be improved as it is a new technology that few people know about. There is a need to spread some awareness about the product while also ensuring that some online forums related to the product need to be updated with a lot of documentation related to the tool.
Some features of runtime don't work well in App Engine. Sometimes we must tune it ourselves or ask Google’s team to improve it. It is the most important thing. Our clients always complain about it.
The documentation is an area that could definitely be improved. Another area of improvement would be the community which is poor and Google could do a lot better on that front. I know that my team members who have been working with the solution for just a year or two find the architecture to be quite difficult.
I would like a simpler deployment tool on laptops. It is a bit complicated at the moment. We know how to do it, but it could be easier to deploy it on laptops.
The only concern is that there is a number of the offerings which are built on their own proprietary technologies. With some of the offerings in Google Cloud, it's difficult to have a path to migrate to other cloud providers.
Everything so far has gone smoothly, I've never had a struggle with the solution. In terms of product improvement, when it comes to billing, I think they should include something that would give the client or user an indication of what's happening so they can be aware of how pricing is being managed.
It would be helpful if they could improve the user interface. It could use some modifications. There needs to be more directions in terms of how to use the solution.
I am limited to sending a photo to five people. I want to be able to send a photo to many people, not just five.
Data consumption of the device could be improved.