CoyoteDB FAQ for high availability Postgres
Short answers to common questions about CoyoteDB: real Postgres, how many computers you need, networks, the witness, coding agents and moving data in and out.
Can I download it?
There is no public download yet.
Is it real Postgres?
Yes. Inside is real PostgreSQL 18. coyote downloads and runs it for you. Anything that speaks standard Postgres can connect.
Do I need to know about databases to use it?
No. The idea is that you install one app, paste an invite code on your other computers, and it stays up. The details on How it works are there if you are curious.
Which computers can I use?
CoyoteDB is built for Windows, macOS and Linux. So far it has been tested on macOS and Linux; Windows has not been tested yet. All the data nodes in one cluster must be the same system type: all Windows, all Linux, or all Mac with the same chip family. The witness and your apps can run on anything. Use computers that stay on; a laptop that sleeps is fine for a single development node, but a poor cluster member.
Do my computers need to be in the same place?
No. They need to reach each other. That can be a plain local network, Tailscale, Twingate, or a UniFi or other site-to-site VPN. Nodes use one encrypted port to talk to each other. It is built for a few locations linked like that, not for spreading one database across the world. See Running Postgres over Tailscale or a VPN.
Where does my data live?
On your machines. CoyoteDB is self-hosted.
How many computers do I need?
One works, with no backup. Two give you a live copy, but no automatic switch-over. Two plus a witness, or three or more, switch over automatically. For most people we recommend two computers plus a witness; when the data really matters, three. See Recommended setups and Node counts explained.
What happens if only one data node is left?
Saves pause for about 30 seconds, then carry on using that node. Until the other node is back and caught up, there is no second copy, and CoyoteDB shows a clear warning. An optional stricter setting makes saves wait instead. See Node counts explained.
What is a witness?
A tiny helper that holds no data. It only votes, so two data nodes can switch over on their own. It can run on any small always-on device, such as a Raspberry Pi, a mini PC or a cheap cloud server. See How automatic failover works.
Do I still need backups?
Yes. Copies between nodes protect you from a dead computer, not from a mistake: a deleted table is deleted on every node at once. See Why you still need backups.
Can my AI coding agent use it?
Yes. It uses CoyoteDB like any Postgres database: give it a normal connection string (coyote conninfo prints one) and a short note on how to use it. See Coding agents.
Can I move my existing Postgres database into CoyoteDB?
Yes, with the standard Postgres tools. Because CoyoteDB runs real PostgreSQL, pg_dump and pg_restore copy a database in from another server, a hosted Postgres service or a laptop. A guided import is not available yet.
Can I move my data out?
Yes, by design. Your data is plain PostgreSQL, so standard Postgres dump tools work on it. There is no lock-in.
Is there a SQL editor, REST API, logins or file storage?
No. CoyoteDB is the database and keeping it up. Use the tools you already use for those.