When comparing ArangoDB vs MarkLogic, the Slant community recommends ArangoDB for most people. In the question“What are the best databases for building social network like apps?” ArangoDB is ranked 9th while MarkLogic is ranked 11th. The most important reason people chose ArangoDB is:
You can model your data as documents or as a graph - no data model lock-in. You can even combine both models and use them in a single AQL query.
Specs
Ranked in these QuestionsQuestion Ranking
Pros
Pro Document and graph-orientend
You can model your data as documents or as a graph - no data model lock-in. You can even combine both models and use them in a single AQL query.
Pro Supports joins
Unlike many NoSQL databases, ArangoDB does support joins in AQL queries.
Pro JavaScript-API
You can extend ArangoDB using JavaScript that runs directly on the Server (Google V8). You can build data-centric microservices that aggregate, validate, transform or enrich data (It's up to you, plain JavaScript) and provide those via a custom API route.
Pro Transaction save
You can use ACID Transactions for short and small data retrieval and/or modification operations in ArangoDB.
Pro Easy cluster setup
Pro Powerful Java Driver (Sync & Async)
ArangoDB has a very good Java Driver for synchronous and asynchronous. In addition the team there is working on a Spring Data integration.
Pro High performance graph database
While known as a document-oriented noSQL database, MarkLogic also provides a standards-compliant triple store that can be used to enrich document-shaped data with semantic links.
Pro Enterprise strength security and transactions
The two fatal flaws of many open source solutions is lack of integrated security at the element level and full ACID transaction support. MarkLogic has both and is trusted with some of the worlds most sensitive data. It is the engine that powers healthcare.gov, which despite some early problems (not caused by MarkLogic) is an amazing technological achievement.
Pro Multi-model database
MarkLogic supports text, documents, key-value/tuple, graphs, tables and object models that don't require extensive data-modeling and normalization that is part of the lifecycle process of relational database management systems. For sites similar to LinkedIn, Facebook, IMDB and even search engines, MarkLogic provides a unique set of features that are all in one box.
Pro No ETL Required
MarkLogic can store all your enterprise data in it's original format without needing to know a schema in advanced. You can shove pretty much any structured or unstructured data directly into MarkLogic, and it will automatically index everything and make it available for future processing. Of course it is fully schema aware and will apply and enforce schema constraints when available, but the tedious normalization that is required for relational databases is not necessary.
Cons
Con Lack of experts
MarkLogic is not as popular as some of it's peers, and even though it is highly standards-based, the pool of talent that is well-versed in the underlying technology is small compared with some competing platforms (e.g, Oracle, IBM, Apache).
Con Enterprise software is not open source and can get expensive. Not for casual projects.
MarkLogic can be downloaded and implemented in development environments for free. However, for production use, it is priced for enterprises, not startups with tight budgets. Open source requires a lot more elbow grease to do the same thing.
