What is the aim of this blog?
The main reason behind creating this blog is to make Presto be more visible. So far it seems that this the first blog which is mainly devoted to this project.
Besides I wanted to share the knowledge about Presto internals. About things that are happening inside while running the query. From my experience I know that understanding how the given tool works under the cover helps me a lot to use that tool. To avoid mistakes, to overcome obstacles, to achieve more.
What will you find here?
I want to post here information about several different topics:
About features of Presto. Not only how to use them and what behaviour to expect, but also share how it was implemented. I hope this will give you a notion when to use what.
You will also find here information like: how to create and install a custom connector, configure kerberos, tune the query. Things that might be documented, but for sure do not have step-by-step explanation.
I will try to find companies and users who successfully run Presto in production. It would be great if they share why they choose Presto, what do are doing with it, where it was great and where it caused pain, what setup do they have etc.?
Engineers and users often live in different worlds. Sometimes engineers get wrong notion about what users need. I would like to overcome that barrier here. I will write about ideas of Presto enhancements, which are not yet implemented (or even not yet planned) to get the early feedback if it is a feature you are waiting for or it is just a waste of time.
Hopefully, benchmarks results will also find a place here. Either these conducted by Teradata or any other 3rd party company. I will try to add an appropriate comment which could explain the reason why numbers look as they look.
Can anything other than a text be found here?
Any utility script or program which I found useful in daily work with Presto and which could be useful for others will be also published here.
Who should read this?
This blog should be read by Presto users. By people who want to broaden the knowledge about Presto, who are using it daily and want to streamline their work.
Who is going to write here?
Mostly me. However, I will try to encourage others to write something also:
colleagues from my team. Presto is a huge project, so it is impossible to be an expert in every part of it. In areas where I do not feel comfortable enough, I will try to ask an expert to give me a hand.
Presto users from 3rd party companies. I will try to collect success stories from actual users.
You? This blog is open-sourced. You can find it source code here. Please do not hesitate to post an issue if you want me to write about something specific or post a pull request if you have something Presto related you want to share with the community. Any of these are very welcome.
Any article written by someone else than me will be specially marked.
Is Teradata (or Facebook) involved in this blog?
I work at Starburst, used to work at Teradata and cooperate with people who are working at Facebook or other companies which are using Presto. However, this blog is a private initiative. Not guided or sponsored by any of these companies.