-
Integrate with new RDS resource support in OpsWorks.
-
Use newly provided search API to find and connect instances in various layers (we currently use the stack configuration directly).
-
Provide some examples using data bags for configuration
-
Native Solr support - use the OS packages to install Solr (instead of buildout and collective.recipe.solr) and allow custom configuration for use with alm.solrindex or collective.solr in the recipe.
-
Automated rolling deployments. We already have this implemented for Plone deployments, where the start/restart of each ZEO client on an instance can be delayed by a fixed amount. It's not clear how to do this for multi-instance deployments where each instance runs single application server. Would probably require running a script (boto?) external to OpsWorks to trigger a deployment on a single instance, monitor it for success and then trigger subsequent deployments.
-
Integration of 3rd party (New Relic, Papertrail, ...) and internal (munin, ganglia, ...) monitoring services.
-
Better documentation - we need feedback about what needs improvement.
-
Instructions for using Docker to test custom recipes.
-
Test suite?