Converting and structuring government NPI entries for direct import into Mongo Collections
Steps for setting up a mongo-backed AEM 6.x cluster for an authoring environment.
AEM is a monolith…a big, big monolith. In addition to being a monolith, or possibly because it’s a monolith, AEM fails to adhere to a n-tier or classic three tier web architecture incorporating a presentation, domain logic, and persistence layers. Traditionally, at least up until the next-gen JCR implementation Oak, AEM bundled all these and more into one, containing: presentation logic internal and external data business logic integrations configurations and secrets In some ways this affords operations engineers and SREs great flexibility and ensures that issues with one publish node can’t affect the others.
A few quick notes on getting AEM 6 off the ground w/ MongoDB (on OS/X, specifically) … Install mongodb via brew: brew install mongodb Unpack the AEM 6 JAR: java -jar cq-quickstart-6.0.0.jar -unpack Take note of the exploded crx-quickstart directory Modify the start script (crx-quickstart/bin/start) providing an additional runmode: crx3mongo Modify the JVM args adding the argument: oak.mongo.uri with the value mongodb://localhost:27017 (update as needed) Start mongod: mongod –dbpath /data/db –httpinterface –journal –directoryperdb –rest Fire up AEM It took my 1st generation MacBook Retina about 4-5 minutes to install and settle.