Claude Code has been a blessing for all our devops work. We use baremetal servers and now have reliable, tested, fault tolerant postgres with streaming replication and barman for backups configured and running, with less than a second of replication lag, including a S3 redundant backup with Yes, it takes some time to set up and test (~3 days in our case, 360GB database). But it's not that complex and the models (Opus 4.8+) know a lot about these days.
PostgreSQL Benchmark: AWS RDS vs. Self-Hosted on Hetzner (2026)
11–16 of 16 posts
Re: PostgreSQL Benchmark: AWS RDS vs. Self-Hosted on Hetzner (2026)
#12> ...That is the realistic "install Postgres and go" baseline. RDS ran on its default parameter group, which AWS tunes for the instance size. Hostim runs its own managed tuning. In other words, the two managed options are tuned out of the box and the self-hosted default is not, and that difference is part of what you are comparing. So... useless comparison vs Hetzner? Akin to comparing two laptops on a compile job, b…
It was "what do you get for least effort and what cost" kind of test
I think it actually favors Hetzner anyway. "Tuned" AWS instance is not impressive comparing it with (still, even after price hike) cheap Hetzner box.
Re: PostgreSQL Benchmark: AWS RDS vs. Self-Hosted on Hetzner (2026)
#13Red flags:
- Tests Postgres 16, not 18.
- Uses t4 burst instances in AWS, shared CPU in Hetzner.
- Test fits in memory.
- Fails to say if load generator is in same AZ.
- Says their own service "includes a standby and automatic failover" but their docs do not describe this feature or its details.
- Says to treat on-demand AWS EC2 price as the floor, does not consider savings plan.
- Their own service describes automatic backups as a potential future feature.
Re: PostgreSQL Benchmark: AWS RDS vs. Self-Hosted on Hetzner (2026)
#14[flagged]
Re: PostgreSQL Benchmark: AWS RDS vs. Self-Hosted on Hetzner (2026)
#15> ...That is the realistic "install Postgres and go" baseline. RDS ran on its default parameter group, which AWS tunes for the instance size. Hostim runs its own managed tuning. In other words, the two managed options are tuned out of the box and the self-hosted default is not, and that difference is part of what you are comparing. So... useless comparison vs Hetzner? Akin to comparing two laptops on a compile job, b…
Yep quite useless. It's well known that postgres defaults are tuned for low resource usage, not for max performance. There are calculators like https://pgtune.leopard.in.ua/ to get a better than default config with next to no effort.
Why can’t Postgres auto-tune these values itself in 2026? Microsoft SQL Server has mostly done a good job of this since about 2005. It will set parallelism thresholds, soak up as much cache RAM and CPU as the runtime environment allows, etc. It will give RAM back when the OS is under memory pressure, and schedule worker threads based on current resources.
A DB with a privileged supervisor process could generally detect how much free RAM, CPU usage/layout, etc at startup and re-evaluate periodically while running.
Re: PostgreSQL Benchmark: AWS RDS vs. Self-Hosted on Hetzner (2026)
#16Red flags: - Tests Postgres 16, not 18. - Uses t4 burst instances in AWS, shared CPU in Hetzner. - Test fits in memory. - Fails to say if load generator is in same AZ. - Says their own service "includes a standby and automatic failover" but their docs do not describe this feature or its details. - Says to treat on-demand AWS EC2 price as the floor, does not consider savings plan. - Their own service describes automat…
There are various differences, but the price differences are incredibly stark and what the article is really about.
Most optimistic AWS bill: $48 list price (and $300 effective price, because neither storage nor network traffic is included in $48) (and, to add insult to injury, AWS is obviously not giving you the same bandwidth as an actual disk, not even half of it, about one third)
Most optimistic Hetzner bill: 19.49 euro (including storage, yes, without backup, including network traffic), 40 with redundancy.
And even though AWS comes with some redundancy by default, for actual redundancy you need a double deployment (or $600 per month)