ClickHouse Reviews

Everyone knows that ClickHouse is the fastest database on the planet, massively reducing the hardware footprint to $100 for most projects. But what makes it shine is the enormous range of procedural functions that can be intertwined with the SQL, which bridge scripting with query negating repetitive queries altogether and allowing itterative, selective and extrapolative access to extended field formats such as complex structures. Finally, the materialised view pre-calculates an abstracted table in the background. All around, it's incredible.

There are no significant drawbacks except a slight learning curve when you get started, but as standard SQL works fine, the transition into more complex queries is painless, and the documentation helps a lot.

Querying billions of records on a $100 device and speed benefit me the most.

Very fast when it comes to querying large amounts of data.

The count operation in specific is super quick.

It is a lot of work if you want to edit any value that is already stored in the database.

We add billions of records to it every day and use different materialized views to gather various analytics.

I like good data compression and fast query execution

I dislike upgrade issues from v.19 to v.21

I would recommend clickhouse for big data analytics

Fast query results over billions of records, storage savings with data compression

Clickhouse is a column-oriented DBMS, and it is fault-tolerant and blazingly fast. Sql like syntax makes it super easy to use. The mathematics using behind scenes using vectors and merging algorithms gives the power to it. I like how quickly we can retrieve the data from the database. In Analytics/Reports we require to get/insert the millions of records and Clickhouse fulfills the purpose. Since it is a column-oriented DBMS, it performs things quickly.

I disliked the fact there I can't make custom functions, I want that in Clickhouse so that we can form our own functions and use them. Materialized Views are still a mystery here they are super valuable but sometimes it tedious to use them and really required multiple joins in materialized views, currently, it's supporting only a single join which is not fulfilling the purpose sometimes. Clickhouse should introduce triggers since materialized views are not helping much here.

If dealing with analytics, then this one is a must-try.

I have been using Clickhouse for reporting in my company. Previously we were using MySQL, and retrieval for half a million was taking 5-10 min there while after using Clickhouse it took only seconds. Clickhouse is really useful when it comes to reporting since there we have to retrieve the huge data and perform transformations, its mysql like syntaxes are super useful that takes no time to adapt this. There are many supportive engines like mergetree, collapsingtree, and engines helping to interact with other dbs like mysql are really useful.

The flexibility of choice of engines for your varying needs

broad coverage of tests

Community support is at times lagging. however test cases are broad

We were solving near realtime aggregation of bus routes for traffic analysis

Searching and partitioning.

Replications.

It takes comapratively more time to start if data present in clickhouse is huge

Searching problems.

Best part of Clickhouse database is it has very easy installation and easy to use for high performance and reliablity

The worst thing in clickhose db is it doesn't support transactions.

I was using for querying the records on large dataset size. It is very fast.