search

2ndQuadrant / pglogical

Logical Replication extension for PostgreSQL 17, 16, 15, 14, 13, 12, 11, 10, 9.6, 9.5, 9.4 (Postgres), providing much faster replication than Slony, Bucardo or Londiste, as well as cross-version upgrades.
http://2ndquadrant.com/en/resources/pglogical/
Other
1.01k stars 153 forks source link

Document pglogical retry behaviour and keepalives #12

Open akretschmer opened 8 years ago

akretschmer commented 8 years ago

it would be fine to have a function for immadiate retry to connect to the provider. After a provider-downtime it takes a few seconds /minutes. And i think it would be fine to see something about the retry-interval within show_subscription_status() , maybe only in down-state. (the timestamp for next automatical retry or something similar)

ringerc commented 8 years ago

I can see the ability to force reconnect would be useful. For now you can just terminate the apply worker.

You can control reconnect intervals with keepalive settings in the connection string.

pglogical doesn't really keep track of reconnect intervals, as such. It only cares about when it restarts workers that have exited, and one possible reason for worker exit is loss of connection.

For info on keepalives see #18 and 362035ef55edaadc0c4ee748061b78d63528131c .

I agree further documentation would be good.