Allow compression for ClickHouse client

[Graphmaxer]

Is your feature request related to a problem? Please describe.
When querying large datasets with a big LIMIT (like 50 000 for example), the time to process the query inside Cube is very slow compared to the same query directly run via ClickHouse CLI in SQL.

Simple SELECT with 50 000 rows on Cube (3.8s):

Same query directly on ClickHouse (instant):

Describe the solution you'd like
The compression option on ClickHouse JS client (https://clickhouse.com/docs/integrations/javascript#compression) can help with large data (https://clickhouse.com/docs/integrations/javascript#tips-for-performance-optimizations).

With compression enabled (1.1s):

Describe alternatives you've considered
I have tried to add more CPU/RAM to the Cube API service, but the issue is still present.

Additional context
Compression cannot be enabled with read only https://clickhouse.com/docs/integrations/javascript#read-only-users

[Graphmaxer]

This works much faster on my side for very large queries with a lot of data.

Example (without cubestore cache)

Before (no compression):

=> More than 12 seconds 🐌

After (with compression):

=> Almost 3 seconds ⚡

[Graphmaxer]

We also got really nice improvements directly in our clickhouse cluster 🥳 Maybe we should set this option to true by default