Some early prototype High Availability features are now available for ambitious installs.

High Availability

Phabricator can now be deployed in a Highly Available cluster configuration, meaning that you can run it across multiple hosts and things will still work if some of them catch on fire. This feature set is still a prototype and some parts of it are currently difficult to configure, but should become more realistic to deploy in the coming weeks. You can find details and progress in T10751.

This service (secure.phabricator.com) has been backed by two separate hosts in different AWS availability zones running fully redundant services for about two weeks (since April 19th). We could lose either node with minimal service disruption and no data loss.

In particular, Phabricator repositories can be configured in a transparent master/master mode where all nodes in a cluster are readable and writable. All upstream repositories (including phabricator/ itself) have switched to multi-master mode. You can find details about this in T4292.

Beyond supporting high availability, cluster hosts (including hosts serving repositories) can also be placed in different geographical regions, although the cluster is not particularly smart about prioritizing nodes that are close to it on the network over more distant nodes yet (see T10884). With a little more work, this will improve performance and availability options for organizations with a global presence.

Repository APIs

Changes are coming to repositories soon to make them fully queryable and editable via the API. You can find details in T10748.

Additionally, these changes will make repositories much more flexible about how they handle the URIs they fetch from, mirror to, and offer to users as clone URIs. You can find details about the URI changes in T10366.

These changes exist today in a new set of "shadow" interfaces, but aren't yet accessible from the UI. We'll cut over to make them authoritative shortly, likely in the next week.