4 ms·

Very cool. How does this compare to something like Microsoft Power Bi? Will I have any issues running it on Linux (Note, im a fairly non-tech user). I have som

Very cool. How does this compare to something like Microsoft Power Bi? Will I have any issues running it on Linux (Note, im a fairly non-tech user). I have some clients that need BI reporting. We've been using Cyfe so far, but it's only good for summaries of data. It's not very good for custom reporting.

salsakran 11y ago
We run it on linux in production and it's the primary platform we recommend for company wide deployments. Power BI is pretty awesome, but pretty complicated when we tried it out. With enough attention it can do very cool things. Our primary audience is folks who don't need the level of customization, and have a lot of non-technical coworkers that want to ask their own questions.
- krmmalik 11y ago
  Brill! Looking forward to trying it
greggyb 11y ago
These are very different product offerings. Power BI offers a mature in-memory columnstore database with a custom functional query language, DAX, with full dimensional modelling facilities. Additionally, if taking advantage of the (freemium) Power BI web portal, you get natural language query and cross-platform mobile apps. This is essentially a SQL query builder on top of a RDBMS with some nice visualizations. I am not trying to disparage the project - in fact I'm often of the opinion that a good query builder is all most organizations need for BI, but the two products in question are vastly different. Power BI has tooling to perform ETL, data modelling, and reporting. Metabase only covers the last piece. In terms of performance, Metabase, being on top of a SQL RDBMS will have the potential to be incredibly speedy, but with large data volumes, you will need a DBA to make sure reporting scales appropriately. With Power BI, that conversation comes much later in the game. 10Ms of rows are trivial with a star schema and sufficient RAM in terms of performance tuning. With a SQL solution, there is a lot more back-end work to tune for an ad-hoc BI workload even at the 1Ms of rows level.

salsakran 11y ago

We run it on linux in production and it's the primary platform we recommend for company wide deployments. Power BI is pretty awesome, but pretty complicated when we tried it out. With enough attention it can do very cool things. Our primary audience is folks who don't need the level of customization, and have a lot of non-technical coworkers that want to ask their own questions.

krmmalik 11y ago
Brill! Looking forward to trying it

greggyb 11y ago

These are very different product offerings. Power BI offers a mature in-memory columnstore database with a custom functional query language, DAX, with full dimensional modelling facilities. Additionally, if taking advantage of the (freemium) Power BI web portal, you get natural language query and cross-platform mobile apps. This is essentially a SQL query builder on top of a RDBMS with some nice visualizations. I am not trying to disparage the project - in fact I'm often of the opinion that a good query builder is all most organizations need for BI, but the two products in question are vastly different. Power BI has tooling to perform ETL, data modelling, and reporting. Metabase only covers the last piece. In terms of performance, Metabase, being on top of a SQL RDBMS will have the potential to be incredibly speedy, but with large data volumes, you will need a DBA to make sure reporting scales appropriately. With Power BI, that conversation comes much later in the game. 10Ms of rows are trivial with a star schema and sufficient RAM in terms of performance tuning. With a SQL solution, there is a lot more back-end work to tune for an ad-hoc BI workload even at the 1Ms of rows level.