When you create an ordered Non-Clustered Columnstore Index in SQL Server 2025, you need to do it single-threaded (MAXDOP = 1) to get a perfectly working Segment Elimination. The same also applies to Ordered Clustered Columnstore Indexes... #sqlserver #columnstore

I've used SQL Server COLUMNSTORE indexes frequently starting around SQL Server 2016/2017.

Whether clustered or nonclustered, columnstore indexes are always stored with compression. By default, it is "COLUMNSTORE" compression. But there's also a COLUMNSTORE_ARCHIVE compression option, which further reduces the size of columnstore data.

#SQLServer
#ColumnStore
#SQL

If you’re in need of a columnar storage engineer, say “hello” to #MariaDB #ColumnStore. 👋

It supports real-time analytics on billions of rows and terabytes of #data. It can’t get any better than that.

Get a formal introduction: https://bit.ly/3fqCkFF

🤩 #MariaDB is truly amazing:
Transactional workloads 👉 #InnoDB
Analytical workloads 👉 #ColumnStore
Read-heavy workloads 👉 #Aria
Write-heavy 👉 #MyRocks
Sharding 👉 #Spider
Scale both reads and writes 👉 #Xpand
Use NoSQL queries or export data to Kafka 👉 #MaxScale

Did you know #MariaDB #ColumnStore is 90% compatible with #InnoDB?

I explain how to use ColumnStore for quick and fast analytics.

📺: http://bit.ly/3FyfXZ9

I did not realize that #Ensembl 's #mysql is built with MyISAM engines as the default.
Wondering if #Aria or #ColumnStore engines are a better choice when building with #MariaDB

Well the #ColumnStore engine, was worth the pain to install in #mariadb
Test file of 10M rows, queries to find the cardinality of the unique values of a specific column ('dim1').
16x & 50x faster than the built-in engines InnoDB & Aria