Running PGSync strictly as WAL consumer

[KylixSerg]

PGSync version: 3.1.0

Postgres version: 15

Elasticsearch/OpenSearch version: 8.14.3

Redis version: 6.2.1

Python version: 3.11

Feature request(consultation):

First off great work here, pgsync for me works well, but for my use case I would prefer to run it in wal replay only, I would like to avoid using Redis and pg events and subsequently the setup of triggers on my tables - reasoning is that I would run pgsync as a cronjob every X minutes therefore the event system is of no use.

Got a working fork, that essentially passes a bunch of flags downstream to skip said stuff, do you see any issues with running pgsync as a strictly WAL consumer, the cronjob will run about say every 20 minutes - my tables does not get large traffic, so I don't risk piling up WAL in that interval.

@toluaina

[toluaina]

Try the polling option. This might satisfy your requirements as it runs a single iteration/cycle and might be more suited for cron

[toluaina]

Just a quick follow-up: you can now run pgsync in polling-only mode. In this mode it ignores triggers and replication logs, continuously polling for changes instead. It’s not exactly a WAL consumer, but it’s pretty close.

pgsync -c schema.json --polling.