1 point by unicorn_startup 10 months ago flag hide 11 comments
peterjarow 10 months ago next
This is really impressive! I'm definitely bookmarking this for further reference. Do you have any plans for implementing new features in PostgreSQL or other databases?
db_admin_notguy 10 months ago next
Good question! We actually have some ideas for enhancing our PostgreSQL configuration and looking into other high-performance databases as well. PostgreSQL is very flexible when it comes to customization, and with the right hands, you can make wonders out of it. Feel free to check our open-source projects on GitHub for more ideas.
anon_user_2020 10 months ago prev next
Users surely appreciate the added convenience! What was your foundation for scaling PostgreSQL like this? Curious to know other details about your setup
sqlexpert15 10 months ago next
It's a parallel-processing system for PostgreSQL, helping us manage massive write-heavy workloads. We'll release more details in a tech blog soon, so stay tuned! (https://ourcompany.tech/postgres-parallel-processing)
thomas0101 10 months ago prev next
@Anon_User_2020, we incorporate the pg_bigm container image provided by Crunchy Data. We customized its settings and applied automated indexing on frequently queried tables with the help of tools like pg_squeeze. In a future post, we are going to share how we trained algorithms to learn from the database and take actions accordingly.
db_guru13 10 months ago prev next
That's fantastic. I am trying to implement a similar approach currently, and this can't be timelier. I'll follow your projects very closely. Quick question: How do you handle failovers and ensuring zero data loss in your setup?
cpit_senior 10 months ago next
@db_guru13, we use Patroni for maintaining PostgreSQL master-slave replication and automating failover. There's a small but powerful microservice that actively monitors the health of our instances, notifying Patroni if a failover is needed. We also utilize WAL-G to manage continuous archiving and backup rotation.
sysadminmaster 10 months ago prev next
What do you think about splitting your monolithic PostgreSQL setup into several shards or even multiple databases? Wouldn't this improve scalability and availability?
pghero_admin 10 months ago next
@sysadminmaster, in our case, we only required around 20 schemas within the single database as we can still achieve high performance and availability. Keeping everything in a single place makes management and maintenance easier.
postgres_addict 10 months ago prev next
Congratulations on your scalability achievements! Do you have any advice for startups and bigger companies that want to follow your successful steps?
databoss 10 months ago next
@postgres_addict, the key takeaway is keeping your infrastructure simple while choosing the right tools for the job. Get to know PostgreSQL's powerful features and capabilities to help you conquer scaling issues. And of course, use monitoring tools extensively to spot bottlenecks and resolve them.