Apache AsterixDB vs Azure DocumentDB

Apache AsterixDB is a free, open source, self-managed Big Data Management System for ingesting, storing, indexing, querying and analyzing data. Its Asterix Data Model extends JSON with object-database concepts, while SQL++ and AQL provide declarative queries for semistructured information. Partitioned LSM-based storage supports B+ tree, R-tree and inverted keyword secondary indexes. External datasets can be queried without loading everything into AsterixDB, using adapters for HDFS and local filesystems. Continuous feeds can ingest sources such as Twitter, RSS and sockets. The HTTP query API accepts SQL++ and can return immediate, deferred or asynchronous results. Installation options include NCService, Ansible and Amazon Web Services. The Hyracks parallel runtime has been scale-tested on up to 1000+ cores and 500+ disks. Documentation describes basic transactional concurrency and recovery, and identifies web data warehousing and social data analysis as suitable uses.

Azure DocumentDB is a document database for applications using BSON and JSON. It supports transactions and vector search, with a maximum document size of 16.0 MB. Deployment options include managed cloud and another deployment model; listed platforms are Linux, macOS, and Windows. The Free Tier is billed Free for the life of the cluster and includes 32 GB of storage and one free database server per subscription. Fixed M-series plans provide different node capacities: M10 starts with 1 vCore and 2 GiB of memory per node, while M300 lists 96 vCores and 384 GiB per node. Between those are configurations with varying memory and compute, including burstable vCores on M25. M200-Autoscale is billed hourly according to usage and offers up to 64 vCores and 256 GiB of memory. Prices for the listed paid tiers are displayed as $-, with actual amounts depending on region and currency. This gives prospective users concrete capacity specifications alongside the free cluster allowance.