How to copy a folder from S3 to Elastic Beanstalk instance on its creation

Question

I'm running a Django web-app on AWS Elastic Beanstalk that needs specific files available to it to run (in fact, an nltk corpus of stopwords). Since instances are created and terminated randomly, I copied the needed folder to the S3 bucket that was created by my elastic beanstalk and planned to add a copy command using awscli to my elastic beanstalk configuration file. But It is not working..

Beanstalk also created an IAM role aws-elasticbeanstalk-ec2-role which is an instance profile that is attached to every instance it launches. This role includes the AWSElasticBeanstalkWebTier policy which seems to grant both read and write access to the S3 bucket:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "BucketAccess",
      "Action": [
        "s3:Get*",
        "s3:List*",
        "s3:PutObject"
      ],
      "Effect": "Allow",
      "Resource": [
        "arn:aws:s3:::elasticbeanstalk-*",
        "arn:aws:s3:::elasticbeanstalk-*/*"
      ]
    }
  ]
} 

I added the following command to .ebextensions/my_app.config:

commands: 01_copy_nltk_data: command: aws s3 cp s3://<my_bucket>/nltk_data /usr/local/share/

But I get the following error when I try to deploy even though I can see the folder in my S3 console

Command failed on instance. Return code: 1 Output: An error occurred (404) when calling the HeadObject operation: Key "nltk_data" does not exist

Any ideas?

Answer 1

I had encountered similar problem. But i found this solution posted by AWS support. My  folder had subfolders and files inside so asw s3 cp command needed --recursive option.

Try it out. It will work.