| We’ve been using Dgraph in production for a few years now. The project is fine for hobby projects but it is NOT production ready. Don’t take my word for it, though… I invite you read through some of the issues reported in their discussion forums and to take a look at their Github contributions over the past year. There was major turmoil in Dgraph Labs (the project’s maintainers) last year which resulted in the CEO and 95% of the engineers exiting the company. They are currently in a rebuilding phase, with limited staff and runway. There are several critical bugs, which lead to either data loss, data corruption or cluster instability, which the current maintainers have failed to fix. Additionally, their customer support is often either unresponsive or unhelpful (even for paying customers). Running a Dgraph cluster is expensive, with heavy memory utilization and favoring vertical scaling. If you need scale, then be prepared to spend big. The documentation is not great and because very few people use this project in production, help is extremely limited. Best of luck to you should you choose Dgraph and to anyone currently using it already. |