PGD-Always-ON v23

Suggest edits

EDB Postgres Distributed 5 in an Always-ON configuration, suitable for use in test and production.

This architecture is valid for use with PGD version 5 only and requires a subscription to EDB Repos 2.0.

Cluster configuration

# Example invocation
[tpa]$ tpaexec configure ~/clusters/pgd-ao \
         --architecture PGD-Always-ON \
         --location-names dc1 dc2 dc3 \
         --active-locations dc1 dc2 \
         --witness-only-location dc3 \
         --data-nodes-per-location 2 \
         --platform aws --instance-type t3.micro \
         --distribution Debian \
         --edbpge 15

# Show all available options
[tpa]$ tpaexec configure --architecture PGD-Always-ON --help

You must specify --architecture PGD-Always-ON.

You must specify a list of location names for the cluster with --location-names dc1 dc2 ….

A location represents an independent data centre that provides a level of redundancy, in whatever way this definition makes sense to your use case. For example, AWS regions, your own data centres, or any other designation to identify where your servers are hosted.

(If you are using TPA to provision an AWS cluster, the locations will be mapped to separate availability zones within the --region you specify. You may specify multiple --regions, but TPA does not currently set up VPC peering to allow instances in different regions to communicate with each other. For a multi-region cluster, you will need to set up VPC peering yourself.)

A PGD-Always-ON cluster comprises a number of locations, preferably odd, each with the same number of data nodes, again preferably odd. If you do not specify any --location-names, the default is to use a single location with three data nodes.

Use --data-nodes-per-location N to specify a different number of data nodes. The minimum number is 2.

If you specify an even number of data nodes per location, TPA will add an extra witness node to each location automatically. This retains the ability to establish reliable consensus while allowing cost savings (a witness has minimal hardware requirements compared to the data nodes).

A cluster with only two locations would entirely lose the ability to establish global consensus if one of the locations were to fail. We recommend adding a third witness-only location (which contains no data nodes, only a witness node, again used to reliably establish consensus). Use --witness-only-location loc to designate one of your locations as a witness. (This location must not be among the --active-locations.)

By default, every data node (in every location) will also run PGD-Proxy for connection routing. To create separate PGD-Proxy instances instead, use --add-proxy-nodes-per-location 3 (or however many proxies you want to add).

By default, TPA will configure PGD-Proxy to use global connection routing, i.e., to elect a write lead from all available data nodes across all locations. You may specify --active-locations l2 l3 to limit connection routing to nodes in the specified locations. This will enable subgroup RAFT and proxy routing for those locations only.

You may optionally specify --database-name dbname to set the name of the database with BDR enabled (default: bdrdb).

You may optionally specify --enable-camo to set the pair of BDR primary instances in each region to be each other's CAMO partners.

You may also specify any of the options described by tpaexec help configure-options.