dremio is awesome, but there is so little document about tech detail when we wanna expand it or we wanna make our little contribute
You’re right - we still have lots to do.
There are tutorials and a few white papers, including one on Dremio’s architecture here: https://www.dremio.com/resources.
Are there specific topics you had in mind? That might help us to prioritize what we add to the documentation.
- firstly, I suggest dremio open its issue page and pull commit page on github.
- provide a performance report about dremio [for example on ten billion raws of data].
- let more users share their experience on dremio.
- provide more document for developers. such as how to write a plugin, how to expand ui.
In fact, we can make a refer to ELK, which has a good eco-system and community culture. in a word, I wish dremio be better and better
Thanks for the great suggestions!
-
I’ll let others comment on this.
-
This is in the works.
-
Can you say more about this? Our hope is that users share their experiences here.
-
We will publish an SDK for sources in a future release. We want this to be easy for the community to use and share with others.
Thanks again!
OK3,
I have been using Dremio at “commercial POC” for past 6 months at major TELECOM customer site. Dremio brings following 3 GREAT abilities to Big Data Analytics, in addition to linear scalability and MP based on DREMEL architecture.
- Federation - Ability to use SQL queries across storages/databases seamlessly.
- Aggregation - Reflections (Accelerations in 1.2). Based on Parquet Columnar storage - VERY well leveraged. Check out EXPLAIN PLANs of SQL queries.
- Data lineage graphs - ABSOLUTELY love this feature. Still need MUCH discipline on Users part!
The product does need to nature, but I strongly believe it is in RIGHT PATH.
These are three MAJOR differentiators and they will take this a long way.
Sold on Dremio!
I strongly agree that the issue page on GitHub should be opened. That makes contribution easier.