4 ms·

Disclaimer: I'm a co-founder of https://logtail.com https://logtail.com, ClickHouse-based hosted log management platform. PostgreSQL, TimescaleDB, and ClickHou

Disclaimer: I'm a co-founder of https://logtail.com https://logtail.com, ClickHouse-based hosted log management platform. PostgreSQL, TimescaleDB, and ClickHouse are all impressive pieces of software. We use both PostgreSQL and ClickHouse at Logtail. ClickHouse shines for true OLAP use-cases and is very hard to beat performance-wise when configured properly. Example: > Poor inserts and much higher disk usage (e.g., 2.7x higher disk usage than TimescaleDB) at small batch sizes (e.g., 100-300 rows/batch). If your consistency requirements allow, you could use the Buffer Table Engine to get blazing fast inserts: https://clickhouse.com/docs/en/engines/table-engines/special/buffer/ https://clickhouse.com/docs/en/engines/table-engines/special... Horizontal scalability and compression are also unbeatable from what I've seen, to name a few. There's a hefty price tag, however: ClickHouse is quite ops heavy and its observability has a seriously steep learning curve. Only go for ClickHouse in production if you really know what you're doing :)

MichaelRazum 5y ago
Could you explain what you mean by ops heavy? Just curious. Actually have a production system where Timescale as well as clickhouse are running in parallel. So far clickhouse didn't do any trouble, but it is rarely used right now.
- eternalban 5y ago
  >> Disclaimer: I'm a co-founder of https://logtail.com https://logtail.com, ClickHouse-based hosted log management platform.
- DevKoala 5y ago
  It is not ops heavy. I believe they are trying to sell you on a managed service with that sentence. It is in your best interest for your team to learn how to maintain ClickHouse themselves. The k8s operator will take you far. By the time you need multi-node clusters you will already have a good idea of how it works. If you need a multi cluster deployment from the start then go with some consulting service I guess, but even then I wouldn’t call it “ops heavy”; the clusters are pretty homogeneous, just a bunch of nodes talking over ZK.
heliodor 5y ago
> 2.7x higher disk usage than TimescaleDB ... A classic tenet of computer science is that you can trade speed for space and vice versa. A database index is an example of this concept. While the article complains about "benchmarketing", its tone is rather dismissive of ClickHouse and parts such as this line caught my eye as unfair.

MichaelRazum 5y ago

Could you explain what you mean by ops heavy? Just curious. Actually have a production system where Timescale as well as clickhouse are running in parallel. So far clickhouse didn't do any trouble, but it is rarely used right now.

eternalban 5y ago
>> Disclaimer: I'm a co-founder of https://logtail.com https://logtail.com, ClickHouse-based hosted log management platform.
DevKoala 5y ago
It is not ops heavy. I believe they are trying to sell you on a managed service with that sentence. It is in your best interest for your team to learn how to maintain ClickHouse themselves. The k8s operator will take you far. By the time you need multi-node clusters you will already have a good idea of how it works. If you need a multi cluster deployment from the start then go with some consulting service I guess, but even then I wouldn’t call it “ops heavy”; the clusters are pretty homogeneous, just a bunch of nodes talking over ZK.

heliodor 5y ago

> 2.7x higher disk usage than TimescaleDB ... A classic tenet of computer science is that you can trade speed for space and vice versa. A database index is an example of this concept. While the article complains about "benchmarketing", its tone is rather dismissive of ClickHouse and parts such as this line caught my eye as unfair.