Replication lag (byte) dashboard panel expects "total_lag" field which is not provided by replication metric

[Timz2009]

Description

The built-in Grafana panel Replication lag (byte) does not display any data because it expects a field that is never collected by the replication metric.

The dashboard query is:

select
  $__timeGroup(time, $agg_interval),
  tag_data->>'application_name'::text as client_info,
  avg((data->>'total_lag')::bigint) as " "
from replication
where dbname in ($dbname)
  and $__timeFilter(time)
  and (data->>'total_lag')::bigint > 0
group by 1, 2
order by 1

However, the replication metric stores:

{
  "sent_lag_b": 0,
  "write_lag_b": 0,
  "flush_lag_b": 0,
  "replay_lag_b": 0,
  ...
}

There is no total_lag field.
Example row:

select data
from replication
limit 1;

returns

{
  "state": "streaming",
  "sent_lag_b": 0,
  "write_lag_b": 0,
  "flush_lag_b": 0,
  "replay_lag_b": 0,
  "write_lag_ms": 1,
  "flush_lag_ms": 1,
  "replay_lag_ms": 2
}

As a result,
(data->>'total_lag')::bigint

always evaluates to NULL and the panel shows No data.

Environment
pgwatch 5.2
PostgreSQL 16.14
Possible solutions

Either:
Update the dashboard to use one of the existing fields (replay_lag_b or GREATEST(sent_lag_b, write_lag_b, flush_lag_b, replay_lag_b)),
or
Add a total_lag field to the replication metric so that it matches the dashboard expectations.

At the moment the dashboard and the collected metric are inconsistent.