humilis-push-processor 0.0.6

Humilis push event processor plugin

Lambda processor for push event sources

A humilis plugin to deploy a Lambda function that processes event notification from an event sources that pushes events to Lambda (e.g. S3 or SNS, as opposed to Kinesis).

Installation

pip install humilis-push-processor

Development

Assuming you have virtualenv installed:

make develop

Configure humilis:

.env/bin/humilis configure --local

You can crate a development deployment (on a deployment stage called DEV) of the Lambda function using:

make create STAGE=DEV

The command above will also create additional resources (such as a S3 bucket) needed to produce a self-contained deployment that you can play with. You can destroy the DEV deployment using:

make delete STAGE=DEV

Testing

To run the local test suite:

make test

To run the integration test suite:

make testi STAGE=[STAGE] DESTROY=[yes|no]

Note that the command above will deploy the processor to the specified deployment stage. If a deployment stage is not specified then TEST will be used. If DESTROY is set to yes all deployed resources will be deleted after tests have completed (this is also the default behaviour if the DESTROY parameter is not provided). You can manually destroy the test infrastructure with:

make delete STAGE=[STAGE]

Deployment secrets

The S3 event processor supports Sentry monitoring out of the box. To activate it you just need to store your Sentry DSN in your local keychain. Using Python’s keyring module:

keyring set humilis-push-processor:[STAGE] sentry.dsn [SENTRY_DSN]

Alternatively you can set environment variable SENTRY_DSN

More information

See humilis documentation.

Contact

If you have questions, bug reports, suggestions, etc. please create an issue on the GitHub project page.

License

This software is licensed under the MIT license

See License file

© 2016 German Gomez-Herrero, FindHotel and others.

Authors

German Gomez-Herrero

original idea and implementation

These people have contributed to humilis-push-processor, in alphabetical order:

• Arnaud Charpentier

• Anatoly Bubenkov

Changelog

0.0.6

• Support for one-to-many mappers

0.0.5

• Mappers do not modify input events directly

0.0.4

• Users can also specify input/output/error streams by name instead of as references to another layer outputs.

0.0.3

• Permission to read from S3

0.0.2

• Fix test suite

0.0.1

• Initial release (germangh)

• Practically a clone of the humilis-kinesis-processor