TimescaleDB continuous aggregate view does not refresh on old buckets

Question

I'm trying to create a continuous aggregate view on a hypertable with the policy to refresh the view once a day.

The first time it runs when I create the view was smooth, everything looked fine on Grafana. However, few days later, I found that the background job mostly failed, hence only partial data was aggregated. I tried to find what happened by using the below query but still have absolutely no idea why the job was failed!

SELECT *
FROM timescaledb_information.job_stats
WHERE hypertable_name = '_materialized_hypertable_17';

All I got was

hypertable_schema	hypertable_name	job_id	last_run_started_at	last_successful_finish	last_run_status	job_status	last_run_duration	next_start	total_runs	total_successes	total_failures
_timescaledb_internal	_materialized_hypertable_17	1013	2021-09-05 10:46:11.810904	2021-09-01 11:53:41.759126	Failed	Scheduled	0 years 0 mons 0 days 0 hours 0 mins 0.022373 secs	2021-09-08 09:38:41.833277	5	2	3

So I tried refreshing the view for those missing buckets. The command was completed within few milliseconds and said the view was already up to date but that was not true! Those partially aggregated buckets were not updated at all.

CALL refresh_continuous_aggregate('daily_system_agg', '2020-09-01', '2020-09-06');

In summary, I got 2 questions:

1. Where can I find more information regarding why the background job was failed?

2. Why the refresh_continuous_aggregate didn't refresh the data?

Script to create continuous aggregate & policy

CREATE MATERIALIZED VIEW daily_requests_agg
            WITH (timescaledb.continuous) AS (
                                             SELECT
                                                 time_bucket('1 day', timestamp) as time,
                                                 gway,
                                                 dest,
                                                 api,
                                                 count(*) as total
                                             FROM requests
                                             GROUP BY
                                                 time,
                                                 gway,
                                                 dest,
                                                 api
                                                 );

SELECT add_continuous_aggregate_policy('daily_requests_agg',
                                       start_offset => INTERVAL '1 month',
                                       end_offset => INTERVAL '1 day',
                                       schedule_interval => INTERVAL '1 day');