aws / elastic-beanstalk-roadmap

AWS Elastic Beanstalk roadmap

Elastic Beanstalk Roadmap

This is the public roadmap for AWS Elastic Beanstalk.

Introduction

This is the experimental public roadmap for AWS Elastic Beanstalk. Knowing about our upcoming products and priorities helps our customers plan. This repository contains information about what we are working on and allows all AWS customers to give direct feedback.

See the roadmap »

Other AWS Public Roadmaps

• AWS App Mesh
• AWS Container services
• CloudFormation coverage
• Amazon EC2 Spot Instances integrations

Security disclosures

If you think you’ve found a potential security issue, please do not post it in the Issues. Instead, please follow the instructions here or email AWS security directly.

FAQs

Q: Why did you build this?

A: We know that our customers are making decisions and plans based on what we are developing, and we want to provide our customers the insights they need to plan.

Q: Why are there no dates on your roadmap?

A: Because job zero is security and operational stability, we can't provide specific target dates for features.

Q: What do the roadmap categories mean?

• Just shipped - Launched and available in roughly the last six months.
• Coming soon - coming up. Think a couple of months out, give or take.
• We're working on it - In progress, but further out. We might still be working through the implementation details, or scoping stuff out.
• Researching - We're thinking about it. This might mean we're still designing, or thinking through how this might work. This is a great phase to send how you want to see something implemented! We'd love to see your use case or design ideas here.

Q: Is everything on the roadmap?

A: Much of our development work for AWS Elastic Beanstalk is included on this roadmap, but not everything. Of course, there will be features we are very excited about that we are going to launch without notice to surprise and delight our customers. We also don’t include every platform security update, but will include significant runtime updates (for example new major versions of Node.js).

Q: How can I request a feature be added to the roadmap?

A: Please open an issue! You can read about how to contribute here. Community submitted issues will be tagged "Proposed" and will be reviewed by the team.

Q: How can I provide feedback or ask for more information?

A: Please open an issue!

License

This library is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License.

To learn more about the Elastic Beanstalk service, head here: https://aws.amazon.com/elasticbeanstalk/

elastic-beanstalk-roadmap's Issues

AWSEBCLI adds additional support for spot instances

AWSEBCLI Version 3.17.0 released on 2019-12-20 and added the following features:

• Added the arguments --on-demand-base-capacity and --on-demand-above-base-capacity. These arguments are used in conjunction with the eb create command.

--on-demand-base-capacity: The minimal number of On-Demand Instances that your Auto Scaling Group provisions before considering Spot Instances as your environment scales up.

--on-demand-above-base-capacity: The percentage of On-Demand Instances as part of additional capacity that your Auto Scaling Group provisions beyond the number of instances specified by the --on-demand-base-capacity option.

• Added the arguments --min-instances and --max-instances. These arguments are also used in conjunction with the eb create command. Here, --min-instances represents the minimal number of Amazon EC2 instances you reqire your environment to have and --max-instances represents the maximum number.

• Added the ability to create a single instance spot environment, e.g. eb create --single --enable-spot.

Windows Server 2019 platform

A new IIS platform version built on top of Windows Server 2019.

.NET Core 3.0 on WIndows Server platforms

We recently launched support for .NET Core 3.0 in our IIS platforms.

ListPlatformBranches support.

See the What's New post at https://aws.amazon.com/about-aws/whats-new/2020/04/aws-elastic-beanstalk-adds-api-support-for-listing-platform-branches/.

Both the Elastic Beanstalk console and the CLI now give you an easy way to determine the LifeCycleState of your environments and both clients now provide notifications that show whether an environment's platform is beta, supported, deprecated or retired. The ListPlatformBranches API documentation can be found at https://docs.aws.amazon.com/elasticbeanstalk/latest/api/API_ListPlatformBranches.html.

Java/Docker Platform Environment Update Issue

I think it's really high priority and it's blocking us from maintaining our environment

https://forums.aws.amazon.com/thread.jspa?threadID=315451

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
For Java and Docker platforms, there is an issue with updating or creating environment variables after the first creation of the environment

Is this request specific to an Elastic Beanstalk platform?
Yes, eu-central-1

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
I want to add/update a variable in Java/Docker platforms. This is currently not possible and gives us an error.

Are you currently working around this issue?
Using the AWS Console and and EBS CLI eb setenv

Additional context
You can check this forum thread for additional information https://forums.aws.amazon.com/thread.jspa?threadID=315451

Attachments
Above ^^ If you need anything else, I'm open for including them from our environment

Corretto on Amazon Linux 2 platform

Beta Corretto Java platforms are now available in all Beanstalk-supported AWS regions. Platform versions for both Corretto 8 and Corretto 11 are available. Check out the release Notes for details.

If you have any comments or questions about the beta platforms, please leave a comment below.

Elastic Beanstalk console enabled the Managed Updates by default

Now when you create an AWS Elastic Beanstalk environment using the Elastic Beanstalk console, managed platform updates are enabled by default whenever possible (on supported platform versions, and with enhanced health enabled), with a weekly update time window that Elastic Beanstalk sets at random. This ensures that your application always runs on a platform with the latest security updates, bug fixes, and minor software stack releases.

https://docs.aws.amazon.com/elasticbeanstalk/latest/relnotes/release-2019-11-25-managed-updates.html

.NET Core on Amazon Linux 2 platform

We are considering building an Elastic Beanstalk platform which allows native support for running .NET Core applications on Amazon Linux 2. Please let us know your interest in this potential platform by reacting with a 👍 on the issue if you want it to be implemented. Additional feedback and comments are also appreciated!

Python on Amazon Linux 2 platform

A Python platform based on Amazon Linux 2 is now available to all customers as a BETA. We'd love it if you could try out the platform and provide any feedback you have. For details about what is different in this platform, check out the documentation page Migrating your Elastic Beanstalk Linux application to Amazon Linux 2.

Feel free to leave comments or complaints in the comments of this issue.

Docker on Amazon Linux 2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

AWSEBCLI on Github

The AWS Elastic Beanstalk Command Line Interface (CLI) is a tool that helps you deploy and manage
your AWS Elastic Beanstalk applications and environments. It has always been our intent to pubilsh the CLI on GitHub and with this effort we hope to achieve a number of goals, to include:

• Encouraging external participation in feature enhancements and bug fixes.
• Fostering an online community of CLI developers (developers helping developers).
• Providing an opportunity to recognize external developers that evangelize the product and contribute toward its improvement.
• Demonstrating more convincingly the idea to current and prospective customers that Beanstalk is builder-friendly.

We look forward to engaging with our avid Elastic Beanstalk developer community!

Elastic Beanstalk support for IMDSv2

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
What do you want us to build?

I want to be able to activate IMDSv2 on Elastic Beanstalk EC2 instances with .ebextensions

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?

IMDSv2 is available for Ec2 instances but not for Elastic Beanstalk

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.

I want Elastic Beanstalk instances to be safe from SSRF using Metadata HTTP API.

Are you currently working around this issue?
How are you currently solving this problem?

I solve this in my code.

Additional context
Anything else we should know?

Attachments
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

PHP 7.4 on Amazon Linux 2 platform (BETA)

PHP 7.4 is now included with Amazon Linux 2 (big thanks to the AL team!). We're planning to include support for it in our upcoming PHP Amazon Linux 2 beta platform.

Elasticache provisioning

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
It'd be nice if it was possible to provision an Elasticache instance within an environment, and have it associated to the environment in the same way an RDS instance is.

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
Most of our applications need some sort of cache server. When provisioning these applications in Elastic Beanstalk, I must create the ElastiCache server separately and manually feed in the connection details via environment variables. This makes it possible to leave Elasticache servers behind when environments are deleted. It's much more of a convenience thing, however if an environment's configuration is "saved" for re-launch/copy, it'd be nice to have the copied environment exactly match the basic configuration of the source, having the cache server managed by EB alongside the database.

Are you currently working around this issue?
I create the environment, then create the Elasticache server separately, and pass in the server's connection details using custom environment variables.

Additional context
Anything else we should know?

Support for HealthcheckType: ELB in aws:autoscaling:asg namespace

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
What do you want us to build?
support for HealthcheckType: ELB in aws:autoscaling:asg namespace

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?
No

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.

The only way today to change to ELB as healthchecktype in elastic beanstalk is with .ebextensions.
It would be awesome if this was available in the elastic beanstalk "namesspaces" so it could be included in cloudformation templates/terraform templates.
The reason for this is that we are a central team to manage a lot of beanstalks for different developers. And it hard to keep track of all the .ebextensions they use.

Are you currently working around this issue?
How are you currently solving this problem?
With .ebextensions.

Additional context
Anything else we should know?

Attachments
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

Glassfish on Amazon Linux 2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

Add support for Spot Instance "interruption notice"

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
Add support for Spot Instance "interruption notice" by automatically launching a new instance (on-demand or spot) when the notice is received instead of waiting until the instance has already been shut down.

Is this request specific to an Elastic Beanstalk platform?
No.

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
This would make it possible to have an Elastic Beanstalk environment with all Spot instances and not worry about them all being shut down at the same time.

Are you currently working around this issue?
Currently we have to set a minimum of 1 on-demand instance to make sure that we don't end up with no instances in an environment.

Additional context
Nothing.

Attachments
No attachments.

Opt-in to Launch Templates even if no features requires it

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
I want a configuration option (like one of these) to opt-in to launch templates, even if I don't use a feature that requires them. This would make some of my ebextension cloudformation template overrides simpler since the launch template will always be there.

Something like this:

option_settings:
  - namespace: aws:elasticbeanstalk:application
    option_name: Use Launch Template
    value: true

Is this request specific to an Elastic Beanstalk platform?
No.

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
Simplify ebextensions that override things in the cloudformation template.

Are you currently working around this issue?
I have some things that is commented out by default, with a comment that it can only be enabled if the deployment is using a launch template.

Additional context
If it matters, my app is open source and I make it easy to deploy on both Heroku and Elastic Beanstalk. Here's the repo: https://github.com/stefansundin/rssbox

Support updating tags via CloudFormation without a replacement

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
ElasticBeanstalk replaces the environment when updating the Tags via CloudFormation. It would be nice if it behaved like the console/api and could do this without a full replacement.

Is this request specific to an Elastic Beanstalk platform?
no

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
When managing environments infrastructure as code is a must, the current implementation makes it very cumbersome to change tags of you do that.

Additional context
see also aws-cloudformation/cloudformation-coverage-roadmap#155

Stop - Hibernate support

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
Support for "Enable hibernation as an additional stop behavior" EC2 launch option.

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
For a forensic security investigation, it may be desirable to have a snapshot of the instance's RAM, which is possible on EC2, but not on Elastic Beanstalk platforms.

Are you currently working around this issue?
Currently you would have to make a decision, whether to leave the compromised instance running and isolated from the network, or power it off. Either is likely to affect its state.

Elastic Beanstalk console major refresh

The Elastic Beanstalk console is getting a major overhaul to its look and feel. In late January, we will release a beta version of the revamped console that will allow customers in us-east-1 to continue using the console as they know it today or optionally explore the new console experience we plan to introduce to all regions later this year. Improvements include streamlined and simplified navigation, a UX that represents a much cleaner and intuitive representation of your Elastic Beanstalk resources, the ability to search and quickly find resources based on a variety of attributes,
quick-links that allow you to seamlessly navigate back to recent environments you touched, and many other new and exicting features!

Elastic Beanstalk Health daemon ignores 4xx errors from Application Load Balancer

Customers can instruct Elastic Beanstalk to ignore 4xx errors that may be thrown by the ALB if you had WAF enabled, for example. This will reduce the false positives from Elastic Beanstalk health checks.

PrivateLink support

We are currently adding support for AWS PrivateLink to Elastic Beanstalk. PrivateLink allows customers to better secure their Elastic Beanstalk environments by ensuring that traffic doesn't traverse the Internet, reducing the exposure to threat vectors such as brute force and distributed denial-of-service attacks. PrivateLink can also provide private connectivity between VPCs, AWS services, and on-premises applications, securely on the Amazon network.

Use latest versions of Puma and Passenger in Ruby platforms

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
The Elastic Beanstalk Ruby platforms have been using very old versions of the Puma and Passenger application servers for several years. The EB environments should be updated to the latest versions of Puma and Passenger, or modified to allow customers to specify the version of Puma and Passenger in the gem file.

Is this request specific to an Elastic Beanstalk platform?
Ruby

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
I would like to use the latest version of Puma or Passenger in applications hosted in Elastic Beanstalk. This isn't possible when the platform installs an older version. These older versions also block new versions of gems that they are dependent on, specifically Rack. Newer versions of Puma and Passenger have new features that are unavailable to customer applications running in Elastic Beanstalk. There is also the issue of missing bug and security fixes included in the newer versions.

Years ago, new Ruby EB platforms did include the current versions of the app servers. This practice stopped a few years ago, and I've been unable to find any public explanation from Amazon as to why. The most logical explanation would be to have the same app server in all Ruby platforms, so use a version that was supported by all platforms. However, at this point in time all currently supported platforms support the latest versions of Puma and Passenger. Another possible explanation would be to avoid breaking customer applications when upgrading Ruby platforms by forcing a switch to a newer app server in a newer platform. However, a customer that needs a specific version of Puma or Passenger can now create a custom Ruby platform or docker container that contains the required version.

For customers that have no preference in app server versions, providing the latest version of Puma or Passenger in the Ruby platforms would seem to make the most sense.

Are you currently working around this issue?
Not at this time. If I need newer versions of Puma, Passenger, or Rack, I wouldn't use the Ruby Elastic Beanstalk platforms. I've not had to do it yet, but I believe that I could create either a custom platform or a docker container that uses the current version of Puma or Passenger, or leaves both app servers out and just includes Rack.

Attachments
Some examples of problems caused by old app server versions:

https://stackoverflow.com/questions/50074609/rails-how-to-get-puma-3-11-for-aws-elasticbeanstalk-application

https://serverfault.com/questions/715116/elastic-beanstalk-with-passenger-standalone-working-configuration

https://forums.aws.amazon.com/thread.jspa?messageID=845581&#845581

Full Django framework support (GIS libraries on Python platform)

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
What do you want us to build?

Make compiled GIS libraries required by Django available in system package repositories so we can install them via EB .config files quickly.

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?

Yes for our use case. Python platform. But the GIS libraries are language-agnostic so it can apply to other platforms as well.

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.

We use Django to build our location-aware mobile apps backend. The builtin geographic features of Django required a few GIS libraries and these libraries are not available in AL repositories so we have to compile from source codes before installation. It takes 20+ minutes to finish the compilation so we have to make a custom EC2 image in advance and a new custom image is required every time a new EB Python platform version is released.

Are you currently working around this issue?
How are you currently solving this problem?

Build a custom EC2 system image:

1. Launch a new EB Python web app.
2. SSH to the EB instance.
3. Install some build tool packages.
4. Compile and install the GIS libraries from source codes (took 20+ minutes).
5. Export path of installed libraries to system environment variables.
6. Save as a custom EC2 system image.
7. Launch new EB instances with the custom EC2 image.
8. Repeat steps above when there is a new EB platform version.

Additional context
Anything else we should know?

RDS PostgreSQL already supports the required PostGIS extension since late 2013.

Many people would appreciate the feature: questions asked on stackoverflow again and again

Attachments
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

No.

Ruby on Amazon Linux 2 platform (BETA)

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

Support for Amazon EC2 Spot Instances

What's New post: https://aws.amazon.com/about-aws/whats-new/2019/11/aws-elastic-beanstalk-adds-support-for-amazon-ec2-spot-instances/

Release Notes: https://docs.aws.amazon.com/elasticbeanstalk/latest/relnotes/release-2019-11-25-spot.html

Remove default security group from EC2-Instance

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
When using aws:autoscaling:launchconfiguration:SecurityGroups (https://docs.aws.amazon.com/elasticbeanstalk/latest/dg/command-options-general.html#command-options-general-autoscalinglaunchconfiguration) the defined groups are added to the EC2-Instance. Nevertheless a default group is also added to the instance. How can this be removed?

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)? multi-container-docker-18.09.9-ce-(generic)

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.
Security hardening.

Are you currently working around this issue?
How are you currently solving this problem?
No workaround.
Edit 2023-02-05: You could try working around by putting the Beanstalk into a VPC with a private subnet (#44 (comment)).

PHP 7.3 on Amazon Linux platform

We recently launched support for PHP 7.3.

Ruby apps should set APP_ENV

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
The Ruby community has decided that RACK_ENV and RAILS_ENV are complicated to use (and used incorrectly a lot). So they decided to add APP_ENV, the one variable to rule them all.

Elastic Beanstalk is setting RACK_ENV=production as a default for new Ruby apps. It should start setting APP_ENV=production as well.

Is this request specific to an Elastic Beanstalk platform?
Ruby

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
It's not hard to set new variables, but better defaults would make things easier for newbies.

Are you currently working around this issue?
I set APP_ENV=production manually after creating an app.

Additional context

• https://github.com/rack/rack/blob/d0218ed43eb4f47bfba614b4671577142dbdbad0/lib/rack/server.rb#L182-L187
• https://github.com/sinatra/sinatra/blob/a4dd24add24f2dd0e7299b9e68e12038138294d3/README.md#environments
• sinatra/sinatra#984

In addition, RACK_ENV=production is not an intended value (as seen in the rack code linked above). It has always been intended to be set to either development, deployment, or none. Unfortunately, Heroku and other websites have been telling Ruby developers to set RACK_ENV and RAILS_ENV to the same value, so this mistake is very common. For APP_ENV, they allow production and other values.

I think the default RACK_ENV value should be changed to deployment. This may have some backwards compatibility concerns, but it is the proper value. One way to deal with this is to also start setting RAILS_ENV=production which should fix this concern for Rails developers.

I also think it would be nice if there was a better way to set default environment variables through .ebextensions somehow. Something like Heroku lets you do via an app.json file (example).

Thanks!

Corretto 8 on Amazon Linux 2 platform (BETA)

What’s New post: https://aws.amazon.com/about-aws/whats-new/2019/11/aws-elastic-beanstalk-launches-public-beta-corretto-al2-platforms/

Release Notes: https://docs.aws.amazon.com/elasticbeanstalk/latest/relnotes/release-2019-11-25-al2-beta.html

Default nginx-gunicorn configuration to use unix sockets in Python on Amazon Linux 2 Platform

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
It would be great for the default nginx-gunicorn configuration to use unix sockets (proxy_pass http://unix:/path/app.sock) as opposed to TCP (proxy_pass http://127.0.0.1:8000;)

Unix sockets are faster in this case. Just imagine the massive path a request will end up proxied from/to now... Cloudfront -> load balancer (ELB) -> nginx (ec2) -> gunicorn (ec2). The TCP connection overhead will be quite interesting with this setup.

Is this request specific to an Elastic Beanstalk platform?
Python

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
This is really just a configuration improvement to boost some performance. It can probably be implemented by hand by adding yet another config file.

Are you currently working around this issue?
Not solving it.

Additional context
ElasticBeanstalk should really provide the "best" possible defaults as possible. Please don't be another Windows.

Lastly, I think the docs around the migration from the previous environment (with Apache) to this new one could be greatly improved.

Attachments

Deploying multiple Java applications in a single Elastic Beanstalk Instance in Asia Pacific (Mumbai) region ap-south-1

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
What do you want us to build?

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?

Yes , Deploying multiple Java applications in a single Elastic Beanstalk Instance in Asia Pacific (Mumbai) region ap-south-1 , its not available in ap-south-1

Are you currently working around this issue?
No

Additional context
Anything else we should know?

Attachments
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

New instance type availability

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
t3a instance types are supported in some regions, but not all. Please deploy support for these instance types to Tokyo!

PHP on Amazon Linux 2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

Namespace available for "Spot Allocation Strategy"

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
What do you want us to build?
Namespace available for "Spot Allocation Strategy".

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?
No

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.

It would be great to have support for "Spot Allocation Strategy" for the ASG namespace.

Are you currently working around this issue?
How are you currently solving this problem?
Just selecting one instance type for spot instances instead of higher ends as well and fail over to ondemand instead.

Additional context
Anything else we should know?

Attachments
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

Support privileged mode on CodeBuild

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
Add support to enable privileged mode on a CodeBuild job created when creating an application version

Is this request specific to an Elastic Beanstalk platform?
All docker-based platforms.

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
It's not currently possible to build docker images for use on EB by the way of the eb cli (I assume due to a lack of support for CodeBuild privileged mode at EB's end).

Are you currently working around this issue?
No. Sort of. As EB doesn't support this currently, a lot of build processes must be created manually (the CodeBuild project, at least). So instead I'm having to use CodePipeline to do a lot of the work for me that I could otherwise (and want to) do with "eb deploy".

Support RDS storage types

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request
Support the ability to provision RDS instances with storage types other than magnetic, and other properties such as monitoring.

Is this request specific to an Elastic Beanstalk platform?
If so, which one(s)?
No

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
It would be nice if there was more support for provisioning RDS instances with more advanced properties than currently available. While not impossible, it just adds a layer of inconvenience to using the platform.

Are you currently working around this issue?
Provisioning an RDS instance via an environment, then manually modifying the created RDS instance with the desired properties.

Additional context
Also, the link for the RDS instance in the "Configuration" tab doesn't work anymore since RDS's console changed to the new design. It no longer filters the console to the instance listed. It would probably make more sense that it takes you to the instance overview page, rather than the list view anyway.

Windows Web Application Migration Assistant

The Windows Web Application Migration Assistant for AWS Elastic Beanstalk is an interactive PowerShell utility that migrates ASP.NET and ASP.NET Core applications from on-premises IIS Windows servers to Elastic Beanstalk.

The migration assistant is available as an open-source project on GitHub, and is able to migrate an entire website and its configuration to Elastic Beanstalk with minimal or no changes to the application. After the assistant migrates the application, Elastic Beanstalk automatically handles the ongoing details of capacity provisioning, load balancing, auto-scaling, application health monitoring, and applying patches and updates to the underlying platform. If you need to also migrate a database associated with your web application, you can separately use AWS Database Migration Service, CloudEndure Migration, or the Windows to Linux Replatforming Assistant for Microsoft SQL Server Databases.

To get started, download the migration assistant from GitHub. Interested parties can also contribute changes and improvements in the form of pull requests.

Node.js on Amazon Linux 2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

Remove public IP on single instance deployments

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Tell us about your request
Currently, when selecting a single instance (no load balancer), and validating that no public is specified. The EB instance is still being created with a Public IP, an ENI. This is not the desired configuration and does not reflect the choice I made as a user.

Is this request specific to an Elastic Beanstalk platform?
All of them.

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.

If I attempt to disassociate the ENI from the instance, it get's re-associated. I can't find the automation in my account so it must be handled by the EB service behind doors.

Are you currently working around this issue?
Not consuming the service and leveraging ECS instead, though not a good alternative due to everything else EB offers. Placing a load balancer on the instance does resolve the problem but there are situations where consumers just want to learn the service and that's when the single instance is ideal.

Additional context
This is coming from an enterprise perspective. Having public instances in private subnets is not the ideal behavior.

Attachments
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)

Python Deploy EB AL2 Fails Constantly

Hi, just wondering if anyone else is having problems with PYTHON 3.7 deployment on EB AL2 as i think I've run into a bug.

Was fighting with it all day yesterday and cannot get apps to deploy properly. Here's what i've tried:

• My App - Dash/flask (failed)
• Example from web - Dash/flask (Failed - Works on AL1 Python 3.6 Environ)
• AL2, new app as a sample app through web interface (Works!)
• Download sample app files, add to zip and deploy:
  • Via EB CLI (failed - won't start)
  • Via EB web (failed - won't start)

If i switch back to the sample it works again. I just cant figure this out. web.service failed to start.
(i'm on eba-cm3rswcy.ap-southeast-2)

It can be reproduced with the Sample app >> Download Sample App > Upload Zip (CLI or Web) > Failure

Help! :)
Thanks

Tomcat on Amazon Linux 2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

Go on Amazon Linux 2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

Managed Platform Updates are failing

Community Note

Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Description:
When users attempt to run managed updates in ElasticBeanstalk environment they are seeing the following error:

=============
Service:AmazonCloudFormation, Message:Template error: instance of Fn::GetAtt references undefined resource AWSEBSecurityGroup.

Cause:
When users run the managed platform update, by default ElasticBeanstalk uses "Immutable updates" with spot instances/custom VPC settings enabled.

Workarounds:

If the environment is created using launch template, the workaround for this problem is to use "On-demand" instances and relaunching the environment with the same application version.

If the environment is created using launch configuration, the workaround would be to use "on-demand" no need of relaunching the environment.

Support running multiple containers using Docker Compose in Docker on AL2 platform

We're working on building out this and other platforms on top of Amazon Linux 2 (AL2). With our AL2 platforms, we are refreshing and rethinking each platform based on feedback we’ve gotten from customers. If you have any questions or suggestions for this platform, please add a comment in the issue.

.NET Core 3.1 on Windows Server platforms

Support for .NET Core 3.1 in our IIS platforms.

Arm-based instance type support

We are considering adding support for Arm based instance types including Graviton instances to Elastic Beanstalk platforms. Arm-based instances provide significant cost savings for compatible applications. Please let us know your interest in this potential improvement by reacting with a 👍 on the issue if you want it to be implemented. Additional feedback and comments also appreciated!

Sharing Application Load Balancers between environments

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave "+1" or "me too" comments. They generate extra noise for issue followers and do not help prioritize the request.

Description
We are researching if we can enable new Elastic Beanstalk environments to share an existing Application Load Balancer (ALB) with host-header and path based routing. This capability will enable ALBs to be shared across different Beanstalk environments and applications within the same VPC.

Feedback requested
We'd love to hear feedback about this potential feature and how you think it should work. In particular we are wondering about a few details:

1. With this feature, Beanstalk would not be involved in creating the ALB for you. You'd create it separately and then reference it in one or more EB environments. Is that the right workflow?
2. You would not be able to switch a Beanstalk environment between a shared ALB and a Beanstalk-managed ALB. Changing would require creating a new environment. Is this a blocker or concern?
3. Would it be valuable for a Beanstalk-managed ALB from one environment to be used as a shared ALB for another environment?
4. Longer term, we want to support sharing of other resources between environments, such as databases. Is this model the right pattern to use?

Use of Amazon EC2 Launch Templates

Until now, the Auto Scaling group in your Elastic Beanstalk environment used a launch configuration attached to the group when it launched Amazon EC2 instances, for example, during scaling actions or during deployments and updates.

Recently we released support for [Spot Instances], and enabling Spot Instances requires using Amazon EC2 launch templates. In the future, additional features might depend on Amazon EC2 launch templates. When you enable such a feature during environment creation or updates, Elastic Beanstalk attempts to configure your environment to use Amazon EC2 launch templates (if the environment isn't using them already). In this case, if your user policy lacks the necessary permissions, environment creation or updates might fail. Therefore, we recommend that you use our managed user policy or add the required permissions to your custom policies. For details about the required permissions, see Creating a Custom User Policy in the AWS Elastic Beanstalk Developer Guide.