rqlite Distributed SQLite Database | Support by cloudimg

This is a repackaged open source software product wherein additional charges apply for cloudimg support services.

rqlite is a lightweight, distributed relational database that uses SQLite as its storage engine and the Raft consensus protocol to replicate writes. You get real relational SQL, transactions and the full SQLite type system, reached over a simple HTTP API that speaks JSON, alongside an interactive command line client. It runs as a single node out of the box, which is a complete working database, and scales out to a fault tolerant Raft cluster when you need one. Because the storage engine is SQLite, a backup is an ordinary SQLite file that any SQLite tool can read. This AMI delivers a fully configured, security hardened rqlite node that answers authenticated queries within minutes of launch, eliminating the manual installation, network hardening and credential setup a bare install requires.

rqlite ships with no authentication by default, permitting every request; this image never runs that way. Authentication is enabled from the first boot with a credential generated uniquely on each instance, so no shared or default database password ever exists in the image, and an unauthenticated call is refused. The HTTP API and the Raft port are bound to loopback rather than exposed on all interfaces, and the instance security group opens SSH only. Where a self managed install co locates database files on the OS disk, this image isolates the entire rqlite data store on a dedicated, independently resizable EBS volume. The rqlite binary is compiled from upstream MIT source with a current toolchain, so fixes that are frozen into an older prebuilt binary are resolved before the image ships.

• rqlite 10.2.7 compiled from the official upstream MIT source at a pinned, provenance-verified tag
• SQLite storage engine with the full SQLite type system and ordinary SQLite backup files
• HTTP API speaking JSON on the loopback interface, with an interactive rqlite command line client
• Raft consensus so a single node grows into a fault tolerant cluster when you need one
• Dedicated data volume for the SQLite database, the Raft log, snapshots and the node identity
• Third party licence notices materialised on the image for every linked component

On first boot a one shot service generates a cryptographically strong password unique to that instance, writes it into the rqlite authentication file readable only by the service account, and mirrors it to a file readable only by root. rqlite refuses to start until first boot has completed, and it re-checks the credential on every start. No shared or default database credential ever exists in the image.

• Amazon EBS - the rqlite data store lives on a dedicated EBS volume; use EBS Snapshots for point in time backups or resize storage independently of the OS disk.
• Amazon VPC - deploy within private subnets and control access with security groups; the database and Raft ports are bound to loopback and never exposed by the image.
• AWS Backup - schedule automated, policy driven backups of the data volume without custom scripting.

Connect over SSH, read the generated password from the root only credentials file, and query the HTTP API over the loopback interface or an SSH tunnel to create tables, insert rows and run SQL. To scale out, join further nodes into a Raft cluster following the guide.

• Lightweight relational store for an application - give a service real SQL, transactions and the SQLite type system without operating a heavyweight database server, reached over a simple JSON HTTP API.
• Fault tolerant configuration or metadata store - start from this single node and grow it into a Raft cluster so critical configuration and metadata survive the loss of a node.
• Edge or embedded relational database - run a small, self contained relational database at the edge or embedded in a deployment, backed up as an ordinary SQLite file any tool can read.

24/7 technical support by email and live chat. Our engineers assist with deployment, connectivity, clustering, backups and database administration. Critical issues receive a one hour average response time.

All product and company names are trademarks or registered trademarks of their respective holders. Use of them does not imply any affiliation with or endorsement by them.