Uplevel pgBadger: The Potential of Postgres for Log Analysis by Henrietta Dombrovskaya

pgBadger is a well-known and widely used tool for identifying performance bottlenecks when using PostgreSQL. But there is even more you can do. Using the newly added "raw output" feature, we built new tools for even deeper insight analysis of the root causes of performance degradation.

1. A ScyllaDB Community
Uplevel pgBadger: The Potential
of Postgres for Log Analysis
Henrietta Dombrovskaya
Database Architect

2. Hettie Dombrovskaya (She/Her)

Database Architect at DRW Holdings
■Prairie Postgres NFP President
■Optimizing everything (not just databases)
■PG Data 2026 Organizer
■Illinois Prairie PUG Organizer
■ACM Chicago Chapter Communications Chair
■LPI Board of Directors

3. pgBadger, why we love it
and why we want more

4. pgBadger report example

5. pgBadger report example

6. Open question

7. Thank you Gills Darold !!!!

8. New pgBadger features
and how we use them

9. New options

--dump-raw-csv – produces parsed log, in a csv format

—csv-separator – assigns separator different from ‘,’

Example:

pgbadger /Users/hettie/chocolate.log –dump-raw-csv –csv-separator # >
chocolate.csv

10. Loading to (another) database

11. How can we use this table

■Session tracing
■Tracing events which happened at the same time
■Performance dynamics
■Precise access control (audit)

12. Example 1: tracing application session

13. Where these BEGIN come from?

15. select
log_id,
log_time,
pid,
duration,
query
from sweets_logs.chocolate_log
where pid =336248
order by log_id
Querying the log

17. Example 2:
“Sometimes, everything is slow!”

22. What else?
Other things already done:
■Complete automation
■Real-time logging
●On the server with ~ 800io/s the storage total usage is ~2GB/hour
■Allow individual logs reload in case of error, preserving the operations order
■Finer time granularity (milliseconds timestamps)
■Associating execution interval with each statement

23. Where to find me
LinkedIn:
GutHub:
hettie-d.github.com
Prairie Postgres:
prairiepostgres.org
PG DATA 2026:
2026.pg-data.org

24. Thank you! Let’s connect.
Hettie Dombrovskaya
[email protected]
@HettieDombr
hdombrovskaya.wordpress.com