Best pratice por stream and index


(Leonardo Ortiz) #1

Hello.

What is the best pratices for index and streams usage?
My general questions is:

  1. Should I have one index per server/service?
  2. One stream for server, service or Event?
  3. Use stream rules or pipelines? Why?
    Can you share you experiences and stream, index and pipelines setup?

(Jan Doberstein) #2

The flexibility is what it makes complicated. Maybe our reworked getting started guide can help you a little: http://docs.graylog.org/en/2.4/pages/getting_started.html

But the way you work and configure your setup is primary based on your usecase. You do not have “one way” to solve.


(Leonardo Ortiz) #3

Fine, will read it a bit more the getting started :slight_smile:

About pipeline x stream rules, its OK use stream rule instead pipeline? I read some users talking that pipeline is more recommended.


(Jan Doberstein) #4

it depends how you have your processing order ( System > Configuration) if that makes sense or not.

You can’t use stream rules you create in processing pipelines when the order of processing have the pipelines after the message filter chain - but you should have the processing pipelines after the message filter chain to have the streamrules already finished.


(system) #5

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.