FoundationDB

FoundationDB Usage Poll


(Rinat Abdullin) #1

What is your primary OS for developing with FoundationDB?

  • Linux
  • Windows
  • OSX

0 voters

Which languages do you use with FoundationDB?

  • Java/Scala
  • .NET/C#/F#
  • Python
  • Golang
  • Node.js
  • Ruby
  • Other

0 voters

Please, respond to these questions by posting a new message:

  1. What are the most exciting FoundationDB features for you?
  2. Which concerns make you (or made you in the past) hesitate moving forward with FoundationDB?
  3. How could the community help you to resolve these issues?
  4. Which problems are you trying to solve with FDB?

[Forum] How would one run a poll here?
(Rinat Abdullin) #2

Answers:

  1. Running multiple data models on the same cluster, fault tolerance
  2. Lack of tooling (esp. visual tools), unusual layer approach that requires getting used to.
  3. Standard cross-platform layers with tooling, more tutorials and documentation.
  4. Delivering HA systems for the retail and finance.

(Rinat Abdullin) #3

Which FoundationDB layers would you like to have?

  • Document store
  • Message queue
  • Background jobs
  • Kafka-like event bus
  • Text search
  • Event store
  • Other (please respond)

0 voters


#4

What is your primary OS for developing with FoundationDB? (Linux, Windows, OSX)

I’v not voted yet, because I use macOS, but every FoundationDB servers and clients are run in docker containers. Should I vote to Linux?

  • What are the most exciting FoundationDB features for you?

Simplicity - the combination of ordered key and transaction.

  • Which concerns make you (or made you in the past) hesitate moving forward with FoundationDB?
  • How could the community help you to resolve these issues?

I struggled about running fdb servers in docker, but I got many help. Thanks!

Currently, I wonder how do I overcome static IP address requirement of fdb cluster nodes as running in Kubernetes (GKE). I don’t have production setup yet though.

Which problems are you trying to solve with FDB?

I’m creating an experimental text search engine. I hope it can be a robust replacement of Elasticsearch in simple use cases.


(Rinat Abdullin) #5

Probably nearly everybody is running FDB in production on the Linux. The question referred to your development environment, so it is OSX.

That is an interesting topic. I might be able to play a little with the K8S soon and would be glad to share any lessons learned.


(Justin Lowery) #6

@hiroshi @abdullin Would something similar to this work?

I haven’t tried that in Kubernetes, though it’s a solution in Docker. You could remove the check for no existing fdb.cluster file, which would edit your fdb.cluster file at startup.

I won’t continue going off-topic here, though I mention it in case it helps, and to find cases where that wouldn’t work, as a few people are trying to pin down a working config for common use-cases with containers.


(Christophe Chevalier) #7

Not everybody :slight_smile: but this will change with .NET Core


(Rinat Abdullin) #8

Sorry, I had to distinguish between running FDB nodes and running the application logic :slight_smile: Probably almost everybody is running FDB processes on Linux, since that is the primary production setup.


(Amirouche) #9

Distributed Ordered Key-Value store that has distributed ACID transaction!

  • It was proprietary.
  • Little or no use outside of Apple
  • Production setup is not good enough
  • Document how to price / size a production fdb cluster

(Amirouche) #10

I mis-read that question. The problem I am trying to solve with FDB is the following:

A single database abstraction primitive to solve our data persistence needs in the domain SEA / lead generation. In particular, I hope it will lower operational cost, make the “on boarding” better and the code more maintainable.