Show HN: BemiDB – Postgres read replica optimized for analytics

github.com

202 points by exAspArk a day ago

Hi HN! We're Evgeny and Arjun, and we’re building a better way to do analytics with Postgres.

We love Postgres for its simplicity, power, and rich ecosystem. But engineers have to still get bogged down with heavyweight and expensive OLAP systems when connecting an analytics data stack.

Postgres is amazing at OLTP queries, but not for OLAP queries (large data scans and aggregations). Even in this case, we’ve still heard from countless scaling startups that they still try to use only a read replica to run analytics workloads since they don’t want to deal with the data engineering complexity of the alternative. This actually works surprising well initially, but starts to break for them as they scale or when integrating multiple data sources. Adding lots of indexes to support analytics also slows down their transactional write performance.

When growing out of “just use Postgres”, companies have to understand and wrangle complex ETL pipelines, CDC processes, and data warehouses — adding layers of complexity that defeat the simplicity that undermines their initial choice for Postgres as their data storage in the first place.

We thought there had to be a better way, so we’re building BemiDB. It’s designed to handle complex analytical queries at scale without the usual overhead. It’s a single binary that automatically syncs with Postgres data and is Postgres-compatible, so it’s like querying standard Postgres and works with all existing tools.

Under the hood, we use Apache Iceberg (with Parquet data files) stored in S3. This allows for bottomless inexpensive storage, compressed data in columnar files, and an open format that guarantees compatibility with other data tools.

We embed DuckDB as the query engine for in-memory analytics that work for complex queries. With efficient columnar storage and vectorized execution, we’re aiming for faster results without heavy infra. BemiDB communicates over the Postgres wire protocol to make all querying Postgres-compatible.

We want to simplify data stacks for companies that use Postgres by reducing complexity (single binary and S3), using non-proprietary data formats (Iceberg open tables), and removing vendor lock-in (open source). We'd love to hear your feedback! What do you think?

banditelol 20 hours ago

Looking at the syncer it seems like copying data to csv from the whole table everytime (?) Code: https://github.com/BemiHQ/BemiDB/blob/6d6689b392ce6192fe521a...

I cant imagine until at what scale can you do this and is there anything better we can do before using debezium to sync the data via cdc?

Edit: add code permalink

oulipo a day ago

Really cool! I have an IoT use-case where I ingest data, I want to keep like the last 3 months or so in Postgresql, and then store the old data as parquet files on S3

I planned initially to do chunks on S3 and do the analytical queries using duckdb, I'm wondering if your tool would be a good replacement?

For now I don't have that many analytical queries, I'm mostly doing visualization of the data points by querying a range (eg last 2 weeks of data for a device)

Does it then make sense to use columnar storage or am I better off with "regular Postgres"?

Or in my case does your approach provide "best of both worlds" in the sense that I could do some occasional analytical queries on past data stored on S3, and regularly access "last 3 months" data for visualization using the data stored in the regular Postgres?

  • jconline543 21 hours ago

    If you just need the Postgres -> S3 archival pattern you described, I built a simpler focused tool: pg-archiver (https://github.com/johnonline35/pg-archiver)

    It:

    - Auto-archives old Postgres data to Parquet files on S3

    - Keeps recent data (default 90 days) in Postgres for fast viz queries

    - Uses year/month partitioning in S3 for basic analytical queries

    - Configures with just PG connection string and S3 bucket

    Currently batch-only archival (no real-time sync yet). Much lighter than running a full analytical DB if you mainly need timeseries visualization with occasional historical analysis.

    Let me know if you try it out!

    • oulipo 17 hours ago

      Really cool! I'll take a look!

      - can you then easily query it with duckdb / clickhouse / something else? What do you use yourself? do you have some tutorial / toy example to check?

      - would it be complicated to have the real-time data be also stored somehow on S3 so it would be "transparent" to do query on historical data which includes day data?

      - what typical "batch data" size makes sense, I guess doing "day batches" might be a bit small and will incurr too many "read" operations (if I have moderate amount of day data), rather than "week batches"? but then the "timelag" increases?

      • jconline543 15 hours ago

        Thanks for your interest! Let me address your questions:

        Querying the data: Yes, you can easily query the Parquet files with DuckDB. The files are stored in a year/month partitioned structure (e.g., year=2024/month=03/iot_data_20240315_143022.parquet), which makes it efficient to query specific time ranges. I personally use DuckDB for ad-hoc analysis since it works great with Parquet. Here's a quick example:

        sqlCopySELECT * FROM read_parquet('s3://my-iot-archive/year=/month=/iot_data_*.parquet') WHERE timestamp BETWEEN '2023-01-01' AND '2023-12-31'

        Real-time data on S3: Currently, the tool is batch-focused. Adding real-time sync would require some architectural changes - either using CDC (Change Data Capture) or implementing a dual-write pattern. I kept it simple for now since most IoT visualization use cases I've seen focus on recent data in Postgres. If you need this feature, I'd be happy to take a look at what you want.

        For data processing the tool works like this:

        It identifies records older than 90 days (configurable retention period)

        Processes these records in batches of 100 (also configurable) to manage memory usage

        Creates Parquet files partitioned by year/month in S3

        Deletes the archived records from Postgres

        The key is that you always have your recent 90 days in Postgres for fast querying, while maintaining older data in a cost-effective S3 storage that you can still query when needed. You can adjust both the retention period and batch size based on your specific needs.

        Let me know if you'd like me to clarify anything or if you have other questions!

        • jconline543 14 hours ago

          Also, for querying both recent and historical data together, you wouldn't need to modify this tool at all. You could just add a separate periodic job (e.g. hourly/daily) that copies recent data to S3:

          sqlCopyCOPY (SELECT * FROM iot_data WHERE timestamp > current_date - interval '90 days')

          TO 's3://bucket/recent/iot_data.parquet' (FORMAT 'parquet')

          Then query everything together in DuckDB:

          sqlCopySELECT * FROM read_parquet([

            's3://bucket/year=*/month=*/iot_data_\*.parquet',  -- archived data
            's3://bucket/recent/iot_data.parquet'             -- recent data
          ])

          Much simpler than implementing real-time sync, and you still get a unified view of all your data for analysis (just with a small delay on recent data).

  • exAspArk a day ago

    Thank you!

    Yes, absolutely!

    1) You could use BemiDB to sync your Postgres data (e.g., partition time-series tables) to S3 in Iceberg format. Iceberg is essentially a "table" abstraction on top of columnar Parquet data files with a schema, history, etc.

    2) If you don't need strong consistency and fine with delayed data (the main trade-off), you can use just BemiDB to query and visualize all data directly from S3. From a query perspective, it's like DuckDB that talks Postgres (wire protocol).

    Feel free to give it a try! And although it's a new project, we plan to keep building and improving it based on user feedback.

    • oulipo a day ago

      Thanks!

      - Can you give me more info about the strong consistency and delayed data, so I can better picture it with a few examples?

      - Also, is it possible to do the sync with the columnar data in "more-or-less real-time" (eg do a NOTIFY on a new write in my IoT events table, and push in the storage?)

      - Would your system also be suited for a kind of "audit-log" data? Eg. if I want to have some kind of audit-table of all the changes in my database, but only want to keep a few weeks worth at hand, and then push the rest on S3, or it doesn't make much sense with that kind of data?

      • exAspArk a day ago

        For now, BemiDB supports only full Postgres table data re-sync. We plan to enable real-time data syncing from Postgres into S3 by using logical replication (CDC), which is much more reliable than PG NOTIFY.

        We use logical replication and this exact approach with our other project related to auditing and storing Postgres data changes https://github.com/BemiHQ/bemi. We're thinking about combining these approaches to leverage scalable and affordable separated storage layer on S3.

        Lmk if that makes sense or if you had any more questions!

        • oulipo a day ago

          Really interesting thanks! I guess my use-case would rather require incremental updates

          Ideally it would just sync in real-time and buffer new data in the Bemi binary (with some WAL-like storage to make sure data is preserved on binary crash/reload), and when it has enough, push them on S3, etc

          Is this the kind of approach you're going to take?

          • exAspArk a day ago

            Yes, we want to use the approach like you described! We'll likely wait until enough changes are accumulated by using 2 configurable thresholds: time (like 30s) and size (like 100MB)

  • Onavo a day ago

    Use a clickhouse FDW or something similar, clickhouse has excellent integration with postgres. They also have a great embedded Python version. Their marketing isn't as good as Duckdb but in terms of stability and performance they are so much better. Duckdb is very very buggy and full of sharp edges but because of their VC funded developer marketing, you don't really hear people talking about it.

    • exAspArk a day ago

      I agree that DuckDB may sometimes be buggy because it's being actively developed with a large surface area (a universal embeddable query engine that works with different storage layers if I were to simplify).

      However, DuckDB (non-profit foundation) != MotherDuck (VC funded). These are two separate organizations with different goals. I see DuckDB as a tool, not as a SaaS or a VC-funded company. My hope is that it'll be adopted by other projects and not associated with just a single for-profit company.

    • thawab 10 hours ago

      Why are you painting Clickhouse as an underdog, with no VC funding? They raised a $250M Series B.

dangoodmanUT a day ago

This is probably the most streamlined/all-inclusive solution out of all that I've seen, but this has definitely been an extremely saturated space in 2024

  • dirtbag__dad a day ago

    What are the other viable players? As mentioned in another thread to this post duckdb is not “production-ready.” That has been a non-starter for us at work.

    • exAspArk a day ago

      That's why our current approach is to build missing or not fully functional features ourselves to move fast. For example, DuckDB performs reads from Iceberg tables not according to the spec, can't perform writes, etc.

  • dangoodmanUT a day ago

    mostly everyone riding on duckdb's tailcoat

dantodor a day ago

Question: is it possible to use BemiDB in ... don't know how to spell it, maybe read-only mode? And by that I mean one Bemi instance that is connected to postgres source, and others that use the produced Iceberg tables to answer queries? Poor man's scalability of the query engine :) ... I would also imagine having an instance that is write-only (reads from postgres and produces the Iceberg tables) and one or many query-only engines. Other than that, great work, will definitely start using it!

  • exAspArk a day ago

    Great ideas! We'll keep this suggestion related to read/write separation in mind. We started with a simple unified solution, but we'll keep iterating, listening and addressing any feedback :)

levkk a day ago

Moving data between systems is problematic. Where this product is actually needed (multi-TB databases under load) is where logical replication won't be able to sync your tables in time. Conversely, small databases where this will work don't really need columnar storage optimizations.

  • woodhull a day ago

    For my use case of something similar on Clickhouse:

    We load data from postgres tables that are used to build Clickhouse Dictionaries (a hash table for JOIN-ish operations).

    The big tables do not arrive via real-time-ish sync from postgres but are bulk-appended using a separate infrastructure.

    • exAspArk a day ago

      Would you be able to share how you implemented "bulk-appended using a separate infrastructure" at a high level?

  • exAspArk a day ago

    Fair point. We think that BemiDB currently can be useful when used with small and medium Postgres databases. Running complex analytics queries on Postgres can work, but it usually requires tuning it and adding indexes tailored to these queries, which may negatively impact the write performance on the OLTP side or may not be possible if these are ad-hoc queries.

    > (multi-TB databases under load) is where logical replication won't be able to sync your tables in time

    I think the ceiling for logical replication (and optimization techniques around it) is quite high. But I wonder what people do when it doesn't work and scale?

    • delive a day ago

      What would you consider to be small or medium? I have a use case for analytics on ~1 billion rows that are about 1TB in postgres. Have you tried on that volume?

      • exAspArk a day ago

        We haven't tested this with 1TB Postgres databases yet, assuming that most companies operating at this scale already built analytics data pipelines :) I'm curious if you currently move the data from this Postgres to somewhere else, or not yet?

        • delive 21 hours ago

          Not yet, mostly just kicked the can down the road due to costs. Like you said in another post, careful indexes on postgres get you quite far, but not nearly as flexible as a columnar DB.

          I think your project is great. I suspect incremental updates will be a big feature for most uptake (one we would need to try this out at least).

nitinreddy88 a day ago

How does update or continuous inserts get written/updated to parquet files? Architecture doesn't show nor anything in docs.

1. All the benchmarks/most of the companies, show one time data exists and try querying/compressing in different formats which is far from reality

2. Do you rewrite parquet data every time new data comes? Or partitioned by something? No examples

3. How does update/delete works. Update might be niche case. But deletion/data retention/truncation is must and I don't see how you support that

  • exAspArk a day ago

    Our initial approach is to do full table re-syncs periodically. Our next step is to enable incremental data syncing by supporting insert/update/delete according to the Iceberg spec. In short, it'd produce "diff" Parquet files and "stitch" them using metadata (enabling time travel queries, schema evolution, etc.)

darkbatman 14 hours ago

We do quite similar but through Debezium/Kafka CDC pipeline to clickhouse. This way primary database is protected. Directly querying from files from postgres might get slower eventually IMO compare.

  • exAspArk 10 hours ago

    This is a great DIY setup. We're hoping to compress this stack and simplify it down to a single binary

leighleighleigh a day ago

Definitely checking this out today! I use postgres for ~30 GB of machine learning data (object detection) and have a couple workflows which go through the Postgres->Parquet->DuckDB processing route.

A couple questions, if you have time:

1. How do you guys handle multi-dimensional arrays? I've had issues with a few postgres-facing interfaces (libraries or middleware) where they believe everything is a 1D array!

2. I saw you are using pg_duckdb/duckdb under the hood. I've had issues calling plain-SQL functions defined on the postgres server, when duckdb is involved. Does BemiDB support them?

Thanks for sharing, and good luck with it!

  • exAspArk a day ago

    Thank you, please give it a try!

    Great questions:

    1. We currently don't support multi-dimensional arrays, but we plan to add support for such complex data structures.

    2. Would you be able to share what type of user-defined functions are these, do they do modify the data or read it?

    • leighleighleigh a day ago

      1. good to hear! 2. The bulk of them are convenience wrappers which resolve UUIDs into other values, so most are read-only with only a single table lookup.

gregw2 a day ago

So it looks like you don't use postgres extensions so you can run this on an EC2 against an Aurora Postgres instance and dump files to S3 Iceberg right?

And can you then have Glue Catalog auto-crawl them and expose them in Athena? Or are they DuckDB-managed Iceberg tables essentially?

  • exAspArk a day ago

    Exactly! You can run it on any server connecting to any Postgres, without installing custom extensions (AWS Aurora supports only a limited number of extensions https://docs.aws.amazon.com/AmazonRDS/latest/AuroraPostgreSQ...).

    The Iceberg tables are created separately from the DuckDB query engine. So you should be able to read these Iceberg tables by using any other Iceberg-compatible tools and services like AWS Athena.

jakozaur a day ago

Cool. Every database or data source (e.g. CRM) should produce Iceberg format for you.

Though a little sceptical of embedding DuckDB. It is easy and better to isolate Read/Write paths, and it has a lot of other benefits.

  • exAspArk a day ago

    Iceberg for the win!

    We actually separate Read/Write paths. BemiDB reads by levering DuckDB as a query engine. And it writes to Iceberg completely separately from DuckDB. I'm curious if that's what you imagined.

    • jakozaur a day ago

      Ideally, I would love many places writing to the Iceberg catalogue and then using it as a data lake.

      In the data lake, I would have a gateway provisioning DuckDB on demand for each user.

      • exAspArk a day ago

        Oh, interesting, thanks for sharing it!

VoxPelli a day ago

The AGPL license is a no-go for me.

While it’s technically true that it’s an OSI license it’s mostly used to scare away competing cloud vendors from hosting the software, which isn’t in spirit of OSS.

Have you looked into the more modern choices?

Like the Business Source License that MariaDB created and uses or the Functional Source License that Sentry created as an improvement over the Business Source License? https://fsl.software/

Both those licenses have a fair source phase that automatically resolves into an open source phase over time.

Thus one gets the best of two worlds: An honest descriptive license for protecting one’s business model + a normal permissive OSS license that ensures longevity and prevents lock-in.

  • senorrib a day ago

    You’re seriously calling out a perfectly valid OSS for not being “in the spirit of OSS”, and pitching for licenses that are explicitly NOT OSS?!

    AGPL couldn’t be more in the spirit of OSS. The entire free software movement started to defend the _users_ freedom, not individual companies’.

    • VoxPelli 16 hours ago

      AGPL is often used by startups to achieve the effect of “fair source” licenses while still being able to claim to be fully OSS.

      AGPL is a poor “fair source” license and a controversial OSS license.

      AGPL in “fair source” projects are always paired with Contributor License Agreements etc that ensures the company behind it owns all the rights to the project and can re-license it however they want without having to abide by AGPL themselves. Which is not in the spirit of OSS at all. And if the company goes out of business, then AGPL makes it really hard for anyone else to eventually pick up and continue on the “fair source” business model of the project.

      Using a proper “fair source” license like the Business Source License or the Functional Source License will make for a better and more honest “fair source” phase of the project and both those licenses will resolve to non-controversial OSS licenses over time.

      So:

      - Is “fair source” in the spirit of “open source”? No

      - Is AGPL an “open source” license? Yes

      - Is AGPL often used to OSS-wash “fair source” projects? Yes

      - Are all AGPL projects “fair source” projects? No, there exists proper OSS-projects with AGPL licenses

      - Is AGPL in startup projects almost always paired with CLA:s that makes the startup play by different rules than everyone else? Yes

      - Is it more in the spirit of “open source” or “fair source” to have the startup play by different rules than everyone else? In the spirit of “fair source”

      - Is AGPL a good choice for “fair source” projects? No, it perpetuates the different rules between the startup and the community forever, even after the startup has ceased to exist, whereas proper “fair source” licenses convert to less extreme OSS licenses over time

      Edit: The only company that I know that does an AGPL project in an “open source” spirit rather than a “fair source” spirit is Sourcehut: https://sourcehut.org/blog/2022-10-09-ip-assignment-or-lack-...

  • exAspArk a day ago

    Our philosophy in general is to go to a more open license over time (vs the other direction). So we might consider other more permissive OSI-approved licenses.

    Would you be able to share why AGPL license is a no-go for you? I'm genuinely curious about your use case. In simple words, it'd require a company to open source their BemiDB code only if they made modifications and were distributing it to other users (allowing modifications and using it internally without any restrictions)

    • senorrib a day ago

      Please, don’t. AGPL is great and you’re fine using it.

    • VoxPelli 16 hours ago

      So you have Contributor License Agreements that enable you to relicense to something like GPL or MIT whenever you want to?

      Because AGPL itself can not be relicensed and it even needed special wording when created to become GPL-compliant.

      Since you’re a startup I believe that you use AGPL to achieve the “fair source” idea (https://fair.io/) – where you yourself can provide a hosted service without providing all your source code while hoping others won’t be as they will need to provide all theirs.

      In simplified terms: It’s an anti-AWS defense. Helping you avoid being outcompeted by a big cloud vendor using your project without paying anything for it.

      And AGPL is a poor “fair source” license, especially as it was never designed to be a “fair source” license but also since it doesn’t give the impression of “fair source” but rather the impression of “open source”, giving an almost deceptive and dishonest look.

      And since AGPL is perpetual, unlike eg BSL and FSL, one is stuck with the “fair source” license forever, rather than having it be converted into a mainstream OSS license over time. It can eg make it hard for someone else to eventually, if your company goes away and the project gets abandoned (which is sadly the most likely outcome of most startups), pick up the project and build a company around that while using similar “fair source” principles like you.

      If you are doing like what SourceHut is doing (https://sourcehut.org/blog/2022-10-09-ip-assignment-or-lack-...) and going all in on AGPL and playing by its rules yourself as well and treating it as “open source” rather than “fair source”, then well done!

      I still would likely want to avoid the legalese complexity of AGPL in my stack though and try to generally stick to permissive licenses and the occasional GPL-licensed projects, like eg Linux. And eg Google has similar guidelines when using OSS-code internally.

  • riiii a day ago

    Because you want to take their hard work, modify it and not share it back to the community?

    I'm not crying that "it's not for you".

    • VoxPelli 16 hours ago

      Absolutely not.

      I want to keep the legalese in my setups at a manageable level, avoid needless lock-ins (AGPL is one of the least compatible licenses out there, code from an AGPL project can only ever be used in another AGPL project) and have it be clear when I contribute to a proprietary project in disguise (which most startup AGPL projects are, thanks to CLA:s) and when I contribute to an actual OSS project

partdavid a day ago

How does this replicate Postgres data? I glanced at the code and saw that it exports to a CSV file then writes out an Iceberg table for an initial snapshot--does it use Postgres logical replication?

  • exAspArk a day ago

    Full table re-syncing is our initial solution. Using Postgres logical replication is next on our roadmap!

    • oulipo 16 hours ago

      So if we have a very large table, it means each time I replicate (how often does that happen?) it needs to rewrite the entire table and upload it to S3, right?

anentropic a day ago

I'm looking for low latency queries over not-very-big data (40-100M rows) in user-facing dashboards

How does the latency of Iceberg-on-S3 compare to say an EBS volume?

  • exAspArk a day ago

    I'd say that querying data from S3 is not ideal when low-latency queries are required. Generally, there could be a few roundtrip requests to fetch metadata (JSON, Avro) and data (Parquet) files, which may lead to around 1s or so latency. However, we have caching on our roadmap (it could be just a simple TTL for the fetched data or some more sophisticated caching depending on the synced & queried data)

    • anentropic 12 hours ago

      How tied is it to S3?

      Could I easily point it to Iceberg tables on another storage target?

      • exAspArk 10 hours ago

        Yes!

        BemiDB natively supports two storage layers, a local disk and S3 (we assumed that most people would choose this in production environments to simplify management).

        When I query Iceberg tables stored on SSD, it works superfast.

gigatexal a day ago

Query Engine: embeds the DuckDB query engine to run analytical queries. Storage Layer: uses the Iceberg table format to store data in columnar compressed Parquet files.

Smart. Imma test this out for sure.

  • arjunlol a day ago

    Thanks! Give it a try and let us know any feedback :)

neeleshs a day ago

Congratulations! I was looking and pg_analytics from ParadeDB hoping this use case would be solved (the dump from pg to parquet part), but it doesnt yet do it.

How does it handle updates?

  • exAspArk a day ago

    Thank you!

    The pg_analytics Postgres extension partially supports different file formats. We bet big on Iceberg open table format, which uses Parquet data files under the hood.

    Our initial approach is to do periodic full table resyncing. The next step is to support incremental Iceberg operations like updates. This will involve creating a new "diff" Parquet file and using the Iceberg metadata to point to the new file version that changes some rows. Later this will enable time travel queries, schema evolution, etc.

nwhnwh 21 hours ago

Any benchmarks against ClickHouse?

  • exAspArk 11 hours ago

    Sorry, we haven't benchmarked it against ClickHouse yet. Our initial point of reference was just Postgres

hoerzu a day ago

Can you give an example if I have 5gig (2 million rows)

How will it be created differently for columnar access?

  • exAspArk a day ago

    We ran some benchmarks (TPC-H, designed for OLAP) with ~10M records https://github.com/BemiHQ/BemiDB#benchmark

    The BemiDB storage layer produced ~300MB columnar Parquet files (with ZSTD compression) vs 1.6GB of data in Postgres.

    • Sesse__ a day ago

      Does TPC-H SF1 really take _one and a half hours_ for you on regular Postgres? Last time I tried (in the form of DBT-3), it was 22 queries and most of them ran in a couple seconds.

      • exAspArk a day ago

        Interesting. I haven't used the DBT-3 kit, does it add any indexes? I manually added these Postgres indexes https://github.com/BemiHQ/BemiDB/blob/main/benchmark/data/cr... to reduce the main bottlenecks on SF0.1 and reduce the total time from 1h23m13s to 1.5s. But SF1 still took more than 1h

        • Sesse__ a day ago

          It adds a bunch of indexes, yes. I don't think anyone really runs TPC-H unindexed unless they are using a database that plain doesn't support it; it wouldn't really give much meaningful information.

          Edit: I seemingly don't have these benchmarks anymore, and I'm not going to re-run them now, but I found a very (_very_) roughly similar SF10 run clocking in around seven minutes total. So that's the order of magnitude I would be expecting, given ten times as much data.

          • exAspArk a day ago

            Got it, thanks for sharing it! We'll try to look into DBT-3 and the indexes it creates to test with SF10

lucasfcosta a day ago

Amazing work, guys! Looking forward to seeing where BemiDB is going.

mrbluecoat a day ago

I hadn't heard of Devbox before, so thanks for sharing

  • exAspArk a day ago

    Haha, it's awesome for isolating project environments (languages, databases, etc.) without using docker

canadiantim a day ago

How does this compare to ParadeDB? Seems to occupy the same space

  • exAspArk a day ago

    We love ParadeDB and their team. Their primary focus is search (Elasticsearch on Postgres), but they also have the pg_analytics Postgres extension (foreign data wrappers and embedded DuckDB).

    The biggest difference is in a Postgres extension vs a separate OLAP process. We want to allow anyone with just Postgres to be able to perform analytics queries without affecting resources in the transactional database, building and installing extensions (might not be possible with some hosting providers), dealing with dependencies and their versions when upgrading Postgres, manually syncing data from Postgres to S3, etc.

winddude a day ago

difference to something like duckdb?

  • simlevesque a day ago

    > > Alternatives

    > DuckDB:

    > - Designed for OLAP use cases. Easy to run with a single binary.

    > - Limited support in the data ecosystem (notebooks, BI tools, etc.). Requires manual data syncing and schema mapping for best performance.

globular-toast a day ago

I don't get how this would do away with the need for some kind of ETL. Most apps use highly normalised schemas that are completely unsuitable for analytical users. Not to mention you wouldn't want to couple your app schema to your warehouse schema. Am I missing something? How would this replace traditional data warehousing?

  • exAspArk a day ago

    Good point. For more complex scenarios, people would still be able to implement, for example, a Medallion Architecture to progressively improve data quality and structure. Because it is Postgres- and Iceberg-compatible (db and data), it's possible to bring more other advanced data tools when it's needed to perform data transformation and movement. Currently, we see it as a Postgres read replica for analytics. But it's easy to imagine that in the future it could be used as a standalone OSS database on top of a data lakehouse with an open format in S3.

woodhull a day ago

As much as DuckDB is cute I've mostly come to believe that Clickhouse is the perfect thing to pair Postgres with. This is especially true now that they've acquired PeerDB and are integrating it into the Clickpipes cloud product.

DuckDB is neat, and I understand why a company like BemiDB would build their product on top of it, but as a prospective customer embedded databases are a weird choice for serious workloads when there are other good open-source solutions like Clickhouse available.

  • maxmcd a day ago

    Using duckdb and apache iceberg means that you can run read replicas without any operational burden. Clickhouse is amazing, but they do not allow you to mount dumb read replicas to object storage (yet).

    I can imagine this product is a very elegant solution for many types of companies/teams/workloads.

    • zX41ZdbW a day ago

      You can mount read replicas on object storage in ClickHouse.

      Example:

          CREATE DATABASE test;
          USE test;
      
          CREATE TABLE hackernews_history UUID '66491946-56e3-4790-a112-d2dc3963e68a'
          (
              `update_time` DateTime DEFAULT now(),
              `id` UInt32,
              `deleted` UInt8,
              `type` Enum8('story' = 1, 'comment' = 2, 'poll' = 3, 'pollopt' = 4, 'job' = 5),
              `by` LowCardinality(String),
              `time` DateTime,
              `text` String,
              `dead` UInt8,
              `parent` UInt32,
              `poll` UInt32,
              `kids` Array(UInt32),
              `url` String,
              `score` Int32,
              `title` String,
              `parts` Array(UInt32),
              `descendants` Int32
          )
          ENGINE = ReplacingMergeTree(update_time)
          ORDER BY id
          SETTINGS disk = disk(readonly = true, type = 's3_plain_rewritable', endpoint = 'https://clicklake-test-2.s3.eu-central-1.amazonaws.com/', use_environment_credentials = false);
      
      And you can try it right now.

      Install ClickHouse:

          curl https://clickhouse.com/ | sh
          ./clickhouse local
      
      Run the query above to attach the table.

      The table is updated in real time. For example, here is your comment:

          :) SELECT * FROM hackernews_history WHERE text LIKE '%Clickhouse is amazing%' ORDER BY update_time \G
      
          Row 1:
          ──────
          update_time: 2024-04-06 16:35:28
          id:          39785472
          deleted:     0
          type:        comment
          by:          mightybyte
          time:        2024-03-21 22:59:20
          text:        I&#x27;ll second this.  Clickhouse is amazing.  I was actually using it today to query some CSV files.  I had to refresh my memory on the syntax so if anyone is interested:<p><pre><code>  clickhouse local -q &quot;SELECT foo, sum(bar) FROM file(&#x27;foobar.csv&#x27;, CSV) GROUP BY foo FORMAT Pretty&quot;
          </code></pre>
          Way easier than opening in Excel and creating a pivot table which was my previous workflow.<p>Here&#x27;s a list of the different input and output formats that it supports.<p><a href="https:&#x2F;&#x2F;clickhouse.com&#x2F;docs&#x2F;en&#x2F;interfaces&#x2F;formats" rel="nofollow">https:&#x2F;&#x2F;clickhouse.com&#x2F;docs&#x2F;en&#x2F;interfaces&#x2F;formats</a>
          dead:        0
          parent:      39784942
          poll:        0
          kids:        [39788575]
          url:         
          score:       0
          title:       
          parts:       []
          descendants: 0
      
          Row 2:
          ──────
          update_time: 2024-04-06 18:07:34
          id:          31334599
          deleted:     0
          type:        comment
          by:          richieartoul
          time:        2022-05-11 00:54:31
          text:        Not really. Clickhouse is amazing, but if you want to run it at massive scale you’ll have to invest a lot into sharding and clustering and all that. Druid is more distributed by default, but doesn’t support as sophisticated of queries as Clickhouse does.<p>Neither Clickhouse nor Druid can hold a candle to what Snowflake can do in terms of query capabilities, as well as the flexibility and richness of their product.<p>That’s just scratching the surface. They’re completely different product categories IMO, although they have a lot of technical &#x2F; architectural overlap depending on how much you squint.<p>Devil is in the details basically.
          dead:        0
          parent:      31334527
          poll:        0
          kids:        [31334736]
          url:         
          score:       0
          title:       
          parts:       []
          descendants: 0
      
          Row 3:
          ──────
          update_time: 2024-11-07 22:29:09
          id:          42081672
          deleted:     0
          type:        comment
          by:          maxmcd
          time:        2024-11-07 22:13:12
          text:        Using duckdb and apache iceberg means that you can run read replicas without any operational burden. Clickhouse is amazing, but they do not allow you to mount dumb read replicas to object storage (yet).<p>I can imagine this product is a very elegant solution for many types of companies&#x2F;teams&#x2F;workloads.
          dead:        0
          parent:      42080385
          poll:        0
          kids:        []
          url:         
          score:       0
          title:       
          parts:       []
          descendants: 0
      
          3 rows in set. Elapsed: 3.981 sec. Processed 42.27 million rows, 14.45 GB (10.62 million rows/s., 3.63 GB/s.)
          Peak memory usage: 579.26 MiB.
      • mdaniel 8 hours ago

        > 3 rows in set. Elapsed: 3.981 sec. Processed 42.27 million rows, 14.45 GB (10.62 million rows/s., 3.63 GB/s.)

        Since you were running $(./clickhouse local) does that mean the query downloaded 14.45GB out of S3 to your machine? The 3.981s seems to imply "no," but I struggle to think what meaning that output would otherwise try to convey

        • zX41ZdbW 2 hours ago

          This is the amount of processed data after decompression. The amount of compressed data is less. You can press whitespace during the query run to see the detailed metrics.

              ReadBufferFromS3Bytes                     8.95 GB
          
          The amount of compressed data read from S3 is 8.95 GB. Note: it sounds quite large, interesting why compression is less than 2x on this dataset. Most likely, it uses just lz4.

          I recommend trying it because it works everywhere. You can run ClickHouse on your laptop, on a VM, or even in Cloudshell in AWS.

      • oulipo 17 hours ago

        When I try your code I get this, any idea?

        Query id: daa202a3-874c-4a68-9e3c-974560ba4624

        Elapsed: 0.092 sec.

        Received exception: Code: 499. DB::Exception: The AWS Access Key Id you provided does not exist in our records. (Code: 23, S3 exception: 'InvalidAccessKeyId'): While processing disk(readonly = true, type = 's3_plain_rewritable', endpoint = 'https://clicklake-test-2.s3.eu-central-1.amazonaws.com/', use_environment_credentials = false). (S3_ERROR)

        • zX41ZdbW 10 hours ago

          It can be a wrong AWS profile in your configuration. In that case, you can correct the configuration, or use an empty server, docker container, or even CloudShell.

      • oulipo 17 hours ago

        Very interesting, can you give more info on how this could be used for instance in my IoT case where I want to keep the last 3 months (say) of data in Postgres, and dump old data in parquet/iceberg on S3, and be able to do analytical queries on the past data? Would that be hard to do?

        And how does the real-time update work? Could I make it so that my latest data is incrementally sync'd on S3 (eg "the last 3-months block" is incrementally updated efficiently each time there is new data) ?

        Do you have example code / setup for this?

        • zX41ZdbW 10 hours ago

          You can store all data in ClickHouse (on S3 or on local storage); there is no need to separate historical and real-time data.

          To insert data into ClickHouse, you use the INSERT query to insert data as frequently as you'd like.

          Alternatively, you can set up continuous replication from Postgres to ClickHouse, which is available in ClickHouse Cloud.

      • maxmcd a day ago

        Whoops, I forgot that tables maintained in the db are not the same as remote archives. :|

        • zX41ZdbW 9 hours ago

          The setup in my example uses a table updated by one ClickHouse server and observed by an infinite amount of read-only replicas, which could run anywhere.

      • nmduc9988 12 hours ago

        how about the case of a clickhouse cluster that use zk and a distributed table

  • exAspArk a day ago

    ClickHouse is definitely a popular choice nowadays. I'm curious whether you self-host ClickHouse or use their Cloud? We wanted to make BemiDB as simple to run as possible with a single binary and object storage (vs large machines, big disks, clustering, running Temporal for CDC, etc.)

hipadev23 a day ago

I don’t totally understand the fascination with storing analytical data on S3. It’s not fast, and if you’re in a write heavy environment it’s definitely not cheap either.

What’s with the avoidance of clickhouse or duckdb paired with insanely fast EBS or even physically attached storage? You can still backup to s3, but using s3 for live analytics queries is missing out on so much of the speed.

  • potamic 21 hours ago

    Would love the authors to pitch in with their use cases, but I think most people simply do not need sub millisecond analytics. This is mostly replacing typical spark pipelines where you're okay with sub second latencies.

    S3 is the cheapest, fully managed storage you can get that can scale infinitely. When you're already archiving to S3, doubling it for analytics saves cost and simplifies data management.

    • hipadev23 5 hours ago

      Sub-millisecond? Literally nobody in large data analysis is doing that. Can you cite some sources or show me what sort of setups (with any DB tech) that you're able to run meaningful queries on S3 sources with even sub-second latency?

  • exAspArk a day ago

    My few cents:

    - Compute and storage separation simplifies managing a system making compute "ephemeral"

    - Compute resources can be scaled separately without worrying about scaling storage

    - Object storage provides much higher durability (99.999999999% on S3) compared to disks

    - Open table formats on S3 become a universal interface in the data space allowing to bring many other data tools if necessary

    - Costs at scale can actually be lower since there is no data transfer cost within the same region. For example, you can check out WarpStream (Kafka on object storage) case studies that claim saving 5-10x

  • nine_k a day ago

    S3 is a protocol understood by "everyone". If you're on AWS, as many are, it's basically the only natural choice. But a number of cloud providers, and a bunch of self-hostable software offer an S3 interface.

    Clickhouse on local NVMe is one possible solution, but then you are married to that solution. An S3 interface is more universal and allows you to mix and match your tools, even though this comes at some expense.

  • ozgrakkurt 15 hours ago

    It is immense amounts of marketing and being lazy to implement actual local storage with replication etc. It just makes everything easier and is marketed a lot.

cocoflunchy a day ago

What I would really love is a dead simple way to: 1) connect to my transactional Postgres db 2) define my materialized views 3) have these views update in realtime 4) query these views with a fast engine

And ideally have the whole thing open source and be able to run it in CI

We tried peerdb + clickhouse but Clickhouse materialized views are not refreshed when joining tables.

Right now we’re back to standard materialized views inside Postgres refreshed once a day but the full refreshes are pretty slow… the operational side is great though, a single db to manage.

paurora a day ago

very cool!! We have the same vision with pg_moooncake: https://github.com/Mooncake-Labs/pg_mooncake/tree/main

From what I understand, the BemiDB experience is akin to PeerDB + Clickhouse. It's not really a Postgres extension?

Glad open table formats are becoming mainstream, for everyone.

  • exAspArk a day ago

    Thanks!

    We love the pg_moooncake extension (and pg_duckdb used under the hood). Although our approaches are slightly different. Long-term, we want to allow anyone to use BemiDB by using native Postgres logical replication without installing any extensions (many Postgres hosting providers impose their restrictions, upgrading versions might be challenging, OLAP queries may affect OLTP performance if within the same database, etc.)

  • shayonj a day ago

    Absolutely stoked for pg_mooncake. I really want to see some of these things happening inside PG and taking advantage of the PG internals + native storage. Only bummer is adoption by places where users are currently, say Aurora. But thats probably a problem for another day :)

    P.S The integration with something like Neon is really cool to see.