I've used Postgres at a few places and the #1 problem was always high availability, not scaling. One Postgres cluster could easily handle 100000 transactions per minute, but when a primary node went down it was a page and manually failing over to the spare then manually replacing the spare. The manual tooling was very finicky but at least it worked, no automated solution came even close. Lack of a good HA story is why I avoid self-managed Postgres as much as possible.
I don't know how the pg scaling story gets fixed unless certain things are rewritten. that's my fear of going all in pg.
mysql has vitess etc & even upgrades are easier. though pg is more extensible.
Couldn't be a better why us :)
Right now I have a project that has very heavy write traffic from multiple services and a web app that reads from this. We are starting to hit the point where no amount of indexing, query optimisation, caching or box upgrades is helping us. We are looking at maybe moving the bulk of the static data to clickhouse to reduce the DB size but I would love to hear if PgDog or other kind of sharding could be useful for this use case.
Not quite. The reason "DBs" like those exist is purely due to fashion. Lets not kid ourselves into thinking they do anything better, save the exception of making data hard to access, which might be a project goal in some cases.
Load balancer with health checks and failover, works out of the box. :) Battle-tested at this point too, so could be worth a look.
This is years of product development with a three person team. If Enterprise sales and support are a big part of your business plan it will suck up a lot more than that.
We sharded over 20 TB that we know about.
This is probably a typo, right? 20TB isn't that big. I would imagine they've sharded a lot more than thatIs there a binary I can run directly?
This kind of tool will help in this case?
Still trying to figure out how this works technically, is the performance gain really just re-write in rust?
If you’re already sharding by tenant for other reasons, OK… But I see CDC to a true OLAP system as more scalable.
PostgreSQL still needs real columnar tables in the core, hopefully one day
https://github.com/pgdogdev/pgdog/commit/36434f93f03dec1d7d4...
I want to have as much fun as the next developer, but that makes me worry, what with supply chain attacks in the news and all.
1. pool exhaustion from idle connections inside open long-running transactions
2. SQLAlchemy's client-side pool using dead connections that PgBouncer had already killed, causing periodic request errors
3. Some tasks have to bypass PgBouncer when they use SET or prepared statements
I've already sharded large datasets at the application layer, but looks like PgDog solves the above problems for any future work?
Also had an issue with it because it cached authentication requests when doing passthrough it seems, I'd changed the roles password, but it kept using the old one, which was no bueno ;).
PgDog seems to make more sense when you really care about a few databases that need massive scale, rather than a simple proxy in front of postgres. I'll keep following the development though, it is much needed in this space, postgres can use all the investment it can get to get it past the single machine scale that it excels at currently.
My question is, has any of them been talked about being upstreamed to postgres itself? Or, adding a custom built in feature to postgres itself?
Edit: It also might be interesting to point out how your solution differs from what the folks at Planetscale are building https://planetscale.com/neki
Surfacing where and how PG is better than Dynamo or any other database is probably a good starting point instead of calling out PG a silver bullet for everything. At the end of the day its all a trade-off.
Wrt. the pooler, how do you compare with pgbouncer?
I'm interested because I have a postgres instance, low-traffic but still like ... tens of r(eads)ps. I was not running anything close to the machine limits but still added pgbouncer to improve performance and didn't see a noticeable difference. I was stress-testing the machine obv., I'm not talking about the 10 rps, lol.
For context, my numbers were something like 10k rps +/- 1k vanilla postgres and like 9k rps +/- 1k with pgbouncer in front of it. So ... slightly slower but big error bars so I wouldn't say for sure. I ended up not using pgbouncer as the benefit was immaterial.
Also yeah, in case you want to check it out, it's the db that backs this project: https://httpstate.com.
This is for DBs that are ~1-1.5TB but doesnt have a huge amount of churn/qps
Effectively what is described here https://www.pgedge.com/blog/always-online-or-bust-zero-downt...
At this point i wonder if i'll ever see that.
If you use something like CloudNativePG they automate parts of the process with cli tools and declarative syntax. Otherwise you take the time to figure it out by hand. It might sound complicated, but just practice on your staging DB, and if all goes well you do the same procedure in prod.
Edit: Apparently Postgres 19 has a patch for one-shot logical replication of sequences! https://www.depesz.com/2025/11/11/waiting-for-postgresql-19-...
1. Control plane to manage multi-node deployments; "works out of the box" experience to make PgDog easy to deploy and use
2. QoS (quality of service): automatically block bad queries from taking down the database
Last but not least, you get SLA-backed support from us (up to P0).
New features are broken down into two categories:
1. Sharding / running Postgres at scale: always open source.
2. Infra management / making it easy to run PgDog at scale: enterprise.
Then again, sharding on a single host probably isn't very useful anyway - but it might work with docker in swarm mode?
Edit:
Performance gains are from having the ability to load balance reads (horizontal scaling for read queries) and scale out writes (with sharding). Once instance bottleneck in Postgres has many faces:
1. Behind schedule vacuums because of too many dead tuples (too many writes)
2. The WALWriter is single-threaded and IO-bound - Postgres can only do about 200-300MB/sec in writes per instance (real prod numbers on EC2 with NVMes and ZFS, basically best case scenario).
3. Bulkheading: single primary is a single point of failure. With 12 primaries, if one fails, 91% of your customers don't notice.
The list goes on. Rust is just a side effect. We love it because it's fast and correct - the perfect match for a database product.
Multi master, from even a conceptual perspective, is incredibly complicated. Databases, transactions, consistency, parallelism are all very complicated.
It’s something that always seems promising at the start but as soon as maintenance and long term improvements enter the picture(ie integrating new Postgres versions), the complexity becomes too much.
In all seriousness, we review every single line of code that goes in and only people who work for PgDog Inc are allowed to merge.
The same old processes vs. threads debate, plus having the ability to scale the coordinator past a single machine. So, if you're OLTP, definitely consider PgDog. OLAP - Citus still wins because of its advanced query engine. We'll get there.
I had to disable application pooling as it was causing read only transactions I could couldnt pin down the cause.
SELECT tenant_id, COUNT(clicks)
FROM users
GROUP BY tenant_id
ORDER BY 2 DESC
LIMIT 25;
Performance is a side effect - definitely needed and we'll do everything we can, but we are not competing with ClickHouse or Snowflake - just trying to make sharded Postgres work with your app.You could also build a watcher side car that watches for changes of the pgdog_users.toml and have pgdog refresh itself then too with this combination. We thought about that but prefer to control the reloads for our needs.
This solves the thousands of clients case for read in a way that is transparent to the clients.
Yes it's required at large scale, especially if you want to distribute reads or shard to a particular geographical area.
We'll get there.
1. Neki as you mentioned 2. PgDog 3. Multigres, headed by original creator of Vitesse
You _could_ make that ACID, but it's not going to be faster than a single machine.
Expanding on that a bit, mongo drivers even have a shared specification of the state machine for monitoring topology changes[1] and algorithm for selecting the server to send an operation to[2] (along with various declarative test cases that the drivers use to validate them alongside the specs in the repo). I think people sometimes underestimate how important the client-side work is to this sort of experience; for all of the faults mongo has had over the years, the amount of investment that they put into the client libraries is something I've never seen anywhere else (although having spent several years working on some of these libraries, my take is likely very biased).
[1]: https://github.com/mongodb/specifications/blob/master/source... [2]: https://github.com/mongodb/specifications/blob/master/source...
its probably the easiest database to run at scale. run & forget. you just have to do a little more work on the data modeling part before you write your application i.e consider your query patterns.
That's exactly right. Get in touch (lev@pgdog.dev), happy to help or at the very least tell you what current works (or doesn't) so you know what your options are.
For both MySQL and PostgreSQL you will need to use some kind of logical upgrades if you want no downtime.
Jun 10th, 2026 Lev Kokotov
Postgres is the only database you need.
The reason DBs like Mongo or Dynamo exist is because Postgres has a scaling problem. If you could make it just work, with 100 TB+ tables and 1M queries per second, we don’t think you would use anything else.
This is why we are building PgDog. Same old Postgres, just with a proxy in front of it, to make it horizontally scalable.
You can deploy PgDog anywhere, including on-prem and in your cloud account: pull our Docker image, change your DATABASE_URL, and make us do the work.
PgDog is serving more than 2M queries per second, in production, across dozens of deployments. We sharded over 20 TB that we know about.
PgDog is open source and anyone can just deploy it, and they do: we have over 1.4M Docker pulls on GitHub.
A new version comes out every week, on Thursdays. Our Discord community is growing. We are there, every day, to answer questions and provide support.
PgDog is a small, three-person startup. So, why use our stuff and trust us with your data?
We are infrastructure engineers, application engineers and generalists. We built apps on Postgres before it was cool and made it work at massive scale.
I ran Postgres at Instacart, where we scaled the company 5x in April of 2020. The biggest problem we had was making Postgres serve 100,000s of grocery delivery orders per minute.
We sharded Postgres on RDS, Aurora and EC2. We fixed the actual problem, using first principles (and a lot of code).
The same technology is now available as an open source product.
Building PgDog is not a pivot. For us, scaling Postgres has been, and is, the only goal.
We built PgDog to run in your cloud, in your colo rack, on-prem, or on your laptop. Wherever you need it, PgDog works, with no dependencies or hidden serverless costs. If you can provide CPUs, our multithreaded code will use them all.
Postgres adoption is only going to increase. With $5.5M from Basis Set, YC, Pioneer Fund and other great investors, we have years of runway, and we are going to make Postgres just work, for everyone, at any scale.
– Lev
P.S. We are building an Enterprise edition of PgDog to make it easier to run in AWS. It comes with SLA-backed support from our team. Give us a call if you want to try it out.
Multi-master is hard. The main issue is what to do with commit/replication lag. It's far "easier" if support for eventual consistency is ok with your use case. In some cases it's not. Also, the problems related to read-only lag can happen on multi-master instances. If somebody does a giant long running query on one of the masters, the target instance needs to hold the data state for the query, even if the underlying DB is getting updates. It also needs to still keep up with other masters. This means the whole cluster can slow down if the multi-master replication is synchronous. Depending on a variety of factors, that can chew up disk space, memory, etc.
There are ways of dealing with these issues (and others), but it comes with tradeoffs with performance, etc.
As long as they don't get undercut by the equivalent of AWS https://aws.amazon.com/rds/proxy/ which is a managed pgbouncer.
Instacart doesn't need "100,000s of grocery delivery orders per minute".
There must be some 0s added for the sake of the story.
Take for example AuroraDB: the sheer engineering it took to make SQL do scalable OLTP at all tells you how much that flexibility actually costs to keep.
> PgDog does not detect primary failure and will not call pg_promote(). It is expected that the databases are managed externally by another tool, like Patroni or AWS RDS, which handle replica promotion.
look, you guys can use a database tool from a guy who thinks he supports replication when the LLM he used to do everything correctly admitted that it doesn't support replication. i won't.
You’d need a ton of faith in these 3 people.
Feels more like it would work better inside of a bigger organization.
The QA tester in me is kinda risk adverse.
1. Let it crash. Increase the RAM, try again.
2. Page to disk (swap), make it slow but ultimately work.
Both have their trade-offs. There is no free lunch here.
It might make 100k row level changes per minute, but that’s a different metric.
https://www.sec.gov/Archives/edgar/data/1579091/000157909126...