*Detailed background on rqlite can be found on [this blog post](http://www.philipotoole.com/replicating-sqlite-using-raft-consensus). Note that master represents 2.0 development (which is still in progress), with a new API. If you want to work with 1.0 rqlite, you can find it [here](https://github.com/otoolep/rqlite/releases/tag/v1.0).*
*rqlite* is a distributed system that provides a replicated SQLite database. rqlite is written in [Go](http://golang.org/) and uses [Raft](http://raftconsensus.github.io/) to achieve consensus across all the instances of the SQLite databases. rqlite ensures that every change made to the database is made to a majority of underlying SQLite files, or none at all.
This starts a rqlite server listening on localhost, port 4001. This single node automatically becomes the leader. To see all available command-line options, execute:
If a node needs to be restarted, perhaps because of failure, don't pass the `-join` option. Using the example nodes above, if node 2 needed to be restarted, do so as follows:
On restart it will rejoin the cluster and apply any changes to its local sqlite database that took place while it was down. Depending on your snapshot threshold, restarts may take a little time. Check out the section below on _Log Compaction_.
rqlite exposes an HTTP API allowing the database to be modified such that the changes are replicated. Queries are also executed using the HTTP API, though the SQLite database could be queried directly. Modifications go through the Raft log, ensuring only changes committed by a quorum of Raft servers are actually executed against the SQLite database. Queries do not go through the Raft log, however, since they do not change the state of the database, and therefore do not need to be captured in the log.
To write data successfully to the database, you must create at least 1 table. To do this, perform a HTTP POST, with a CREATE TABLE SQL command encapsulated in a JSON array, in the body of the request. For example:
where `curl` is the [well known command-line tool](http://curl.haxx.se/). Passing `-L` to `curl` ensures the command will follow any redirect (HTTP status code 307) to the leader, if the node running on port 4001 is not the leader. It is *very important* to set the `Content-Type` header as shown.
The use of the URL param `pretty` is optional, and results in pretty-printed JSON responses. `explain` is also optional. If included, the response will include some basic information about the processing that took place -- how long it took, for example.
Transactions are supported. To execute statements within a transaction, add `transaction` to the URL. An example of the above operation executed within a transaction is shown below.
When a transaction takes place either both statements will succeed, or neither. Performance is *much, much* better if multiple SQL INSERTs or UPDATEs are executed via a transaction.
The behaviour of rqlite when more than 1 query is passed via `q` is undefined. If you want to execute more than one query per HTTP request, place the queries in the body of the request, as a JSON array. For example:
Another approach is to read the database file directly via `sqlite3`, the command-line tool that comes with SQLite. As long as you can be sure the file you access is under the leader, the records returned will be accurate and up-to-date.
rqlite replicates SQLite for fault-tolerance. It does not replicate it for performance. In fact performance is reduced somewhat due to the network round-trips.
Depending on your machine, individual INSERT performance could be anything from 1 operation per second to more than 10 operations per second. However, by using transactions, throughput will increase significantly, often by 2 orders of magnitude. This speed-up is due to the way SQLite works. So for high throughput, execute as many operations as possible within a single transaction.
An Administration API exists, which dumps some basic diagnostic and statistical information, as well as basic information about the underlying Raft node. Assuming rqlite is started with default settings, the endpoints are available like so:
rqlite does perform log compaction. After a configurable number of changes to the log, rqlite snapshots the SQLite database. And at start-up rqlite loads any existing snapshot.
This is new software, so it goes without saying it has bugs. It's by no means finished -- issues are now being tracked, and I plan to develop this project further. Pull requests are also welcome.
rqlite reports a small amount anonymous data to [Loggly](http://www.loggly.com), each time it is launched. This data is just the host operating system and system architecture and is only used to track the number of rqlite deployments. Reporting can be disabled by passing `-noreport=true` to rqlite at launch time.
This project uses the [Hashicorp](https://github.com/hashicorp/raft) implementation of the Raft consensus protocol, and was inspired by the [raftd](https://github.com/goraft/raftd) reference implementation. rqlite also uses [go-sqlite3](http://godoc.org/github.com/mattn/go-sqlite3) to talk to the SQLite database.
