Graylog Enterprise production infrastructure setup



Our company decided to implement a graylog based log storage system for few applications. We want to buy the enterprise version. For the start we would like to create a small system of one machine server+mongodb and three vms for elasticsearch cluster (id also like to know whether this is enough). We seek some advice for assessment for the servers power.

Data assumptions:

250 GB /day
up to 10000 lines/s
about 50 sources
two weeks retention

rather simple queries for monitoring and error analysis purposes


I’d like to know the recommended CPU Memory Disk for this setup, maybe the number of nodes.

We are not focused on HA for graylog server and mongo. Rather looking for a fast and easy way to store fresh log.

(Jan Doberstein) #2

you will hold ~3500 GB Logs and want to have them searchable. You might run into some issues with only 3 elasticsearch nodes.

Additional if the Graylog Server has enough resources highly depends on the usage, that means how intense you work with the log data. How many processing you run during ingest. How many alarms/streams/dashboards you use.

Scaling is easy, just start and prepare to add additional resources.

(system) closed #3

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