/ses-mailer

Python-based (Boto) mailer for AWS Simple Email Service (SES)

Primary LanguagePythonApache License 2.0Apache-2.0

SES Mailer

Python-based (Boto) mailer for AWS Simple Email Service (SES). Demonstrates how to send email faster using AWS SES without a backend.

Based on the excellent code from AWS at https://github.com/awslabs/aws-support-tools/tree/119b4a6c22472515008a049506d6dbfcdee59ac8/SES/SESMailer, with the following features added:

  • Recipient email address validation
  • An attempt at idempotence, so that multiple successive runs of the script do not result in sending duplicate messages to recipients
    • Suitable for invocation by AWS Lambda
  • Displays statistics at script conclusion
    • Number of messages sent
    • Number of message resends skipped (due to idempotence)
  • Improved error/exception handling and reporting
    • Print all errors, instead of uploading to S3
  • Improved cleanup of inputs from S3 bucket, to prevent accidental future sending mistakes
  • Simplified execution flow when using a single thread
  • Various bugfixes

How it works

SES Mailer uses AWS Lambda to send mass email. It is invoked by S3 PUT event when a mailing list is dropped into the designated bucket. The function takes its message content from a HTML and plain text message file dropped in the same bucket. It sends multipart/alternative formatted emails.

Usage

  1. Deploy ses_mailer.py on AWS Lambda with a timeout setting of 5 minutes. Make sure the lambda Role has S3 read/write permissions to the bucket and ses:SendRawEmail permission. Configure Lambda environment variables:

    • REGION=us-east-1
    • MAX_THREADS=1
    • TEXT_MESSAGE_FILE=email_body.txt
    • HTML_MESSAGE_FILE=email_body.html

  2. Create a S3 bucket and set put event to trigger this lambda function.

  3. In the S3 events configuration, set the event suffix to .gz.

  4. Write your HTML-formatted email in file email_body.html and upload to S3 bucket.

  5. Write the plain-text version of your email message in file email_body.txt and upload to S3 bucket.

  6. Create a mailing list file (e.g. mailing_list_14032016.csv) with contents in the below csv format. Don't include first row as header fields.

    Sender Name <me@example.com>, Recipient Name <you@example.com>, subject

  7. Compress the file using gzip. e.g. gzip -kf mailing_list_14032016.csv creates mailing_list_14032016.csv.gz

  8. Upload the gzipped file mailing_list_14032016.csv.gz to the S3 bucket and it will trigger this lambda function.

  9. This function will start sending email to all addresses in the csv file and log failures to the console

Tip: You can send even faster by splitting email list into multiple smaller csv files when the number of addresses exceed over a few 100,000s or increase MAX_THREADS environment variable value to something higher depending on your SES TPS limit.