How Can I Easily Add Environment Variables To Multiple Lambda Functions?

user_8193858 0
Viewed 2682

I'm working on an AWS serverless application, I'm new to AWS so please correct any misconceptions in this question. We have around 30 lambda functions doing basic CRUD operations on a SQL database hosted in RDS. We used EntityFramework Core to create the database. This means that, in order to use the database, the lambda functions each need access to the EF connection string. I didn't want to leave the connection string in memory (it contains a plaintext password) so I put it in an encrypted environment variable.

The only way I could work out how to do this was to add an encrypted environment variable through the lambda management console GUI to every lambda function individually. This was tedious but it worked. I've now changed the solution a bit and we require a different set of lambda functions, these need environment variables adding to them too and I don't want to do it manually again.

My question:

Is there an easy way to add encrypted environment variables (or something similar) that can be accessed by all of my lambda functions? Is there a method to add them in bulk? A higher-level variable that I can use?

I have tried to find information in the Amazon docs (here for example) but had little success.

aws-lambda .net-core amazon-web-services environment-variables c#
4 Answers

Lambda is only concerned about getting the environment variables that it needs and it is not responsible for centrally managing environment variables.

What you would need to do is to handle your environment variables in your deployment process.

How do you currently deploy your 30 lambda functions?

If you use CodeBuild, it can access AWS EC2 Parameter Store where you can centrally store and manage your sensitive environment variables.

Other CI/CD tools also have their own way of centrally managing environment variables.

Update:

You can use the AWS EC2 Parameter Store programmatically. This means that you can retrieve your stored environment variables from inside your Lambda function during startup so you don't need to do this at deployment time.

Doing it on deployment time is still better if that is possible in your use case.

user_12526470

I would not call this as a solution rather a work around to address your problem. You can configure your environment variables and store it onto S3 bucket. And using S3 client within your each lambda , you can read specific variable you are interested in. Hope this helps.

user_51442680

I would suggest to add ConnectionString at AWS console and to not use production connection string at config file to avoid those kinds of problems. Log into AWS console, go to lambda function and add connection string as an environment variable. PS. In case you will struggle with syntax. If your json config file looks like this:

{
  "Settings":
    {
      "ConnectionString":"",
      "SomethingElse"
  }
}

Use this syntax to configure connection string:

Settings__ConnectionString

user_70850700

I am late to the party. How about saving your connection string into an DynamoDB table and let each lambda take from it. I assume one connection string for all of it. Most of the time for SQL Server.

user_125533940
Related