Bugs

Removes lag checks when database is a primary database.

Bug fixes

No longer exits process with status 1 when getting an error querying for node info. Removed superfluous writing to header issue which was creating a lot of noise.

Support Postgres 13

Adding columns wal_status, safe_wal_size, written_lsn, flushed_lsn, sender_host, and sender_port to support postgres 13.

Note: This release includes commits ef2a564, d206e84, 78ae621, and 2d1cd67.

Support multi hop

Previously, pgreba has been limited to health checking replicas only 1 hop from the primary. Now you can configure to health check replicas n number of hops by setting max_hop value in the yaml config file.

Allow pgreba to start when postgres is down

No longer creating a db connection in initialization. This allows pgreba to start even when postgres is down or when a db service hasn't started yet, and for pgreba to run properly when it does come back up.

Measure byte lag using replay lag

We were measuring receive lsn and now we're measuring replay lsn.

Cached data source, db pool timeout added, and bug fix

Improvements

• Added cached data source to reduce number of connections opened to the database
• Fixes "broken pipe" issue by adding a pool connection timeout

Bugfix

• No more lingering replica db pool connection left open.

PRs

#13
#14

Optional max_allowable_byte_lag param

Configure max_allowable_byte_lag with query param like so:

GET /replica?max_allowable_byte_lag=12344

v1.1.0

Works with postgres 10!

v1.0.0

Merge pull request #8 from film42/schinn/version_flag

added version flag