Hey folks, I've been working a way for small teams...
# all-things-deployment
a
Hey folks, I've been working a way for small teams with small budgets, non-profits, academia etc. to deploy DataHub in as simple a way as possible. If you're in that boat, I've set up a guide: https://github.com/languageconvo/datahub-deployed Some quick notes: • Read the readme 🙂 You should probably use Acryl Data's managed product when it's available! • Cost is ~ $150 - $180/month although you might be able to get away with ~$100/month • Non-kubernetes. Having devops experience will be needed though, we use AWS RDS, Elasticsearch, and docker-compose quickstart • Would love feedback from anyone that has time and knows what they're doing (as opposed to us/this being our first time every deploying DataHub) -- is what we come up with an OK long-term strategy, or are we going to run into major problems? Thanks!
datahubbbb 3
🙌 6
thanks bear 2
A few other notes: • We started with DataHub's Docker quickstart
docker-compose-without-neo4j.yml
, removed MySQL and Elasticsearch, set those two up as separate managed services leaving the rest of DataHub's front and back end in the docker-compose. • Updating to new DataHub versions should hopefully be fairly easy, just involves a few minutes of getting the new docker-compose file and uploading it to AWS
b
Awesome work @astonishing-lizard-90580 - thank you so much for sharing!!
c
Thanks @astonishing-lizard-90580, this is really helpful!
t
plus1 Thanks for sharing! Came here for some advice on creating a small production deployment and this looks very doable 🙂