Graylog AWS Load Balancer


(Starz31) #1

Hi,

Trying to setup Graylog as follows:

Inside VPC,
Graylog 2.3.2-1 as ECS container (Non reserved host port mapped to container port 9000)
AWS Application Load Balancer with listener and target group for routing (host based routing)
EC2 MongoDB instance
AWS Elasticsearch

Config values overwritten
GRAYLOG_WEB_ENDPOINT_URI = https://dnsname.domain.com/api
GRAYLOG_ELASTICSEARCH_HOSTS = Elasticsearch Endpoint
GRAYLOG_MONGODB_URI = Mongodb Uri

From the logs, I can see Graylog node in ECS connecting to ES and MongoDB, however it gives ‘503 Service Temporarily unavailable’ when I browse to Graylog url?

Do more config changes need to be done? Should I be using AWS Network load balancer instead?


(Jochen) #2

What’s the complete configuration of your AWS ALB?
What’s in the logs of the Graylog Docker container?


(Starz31) #3

ALB Config

  1. Internet facing
  2. Listening on HTTP port 80, host based routing set to forward traffic to Graylog target group
  3. Security group to accept HTTP port 80 traffic

Target Group Config

  1. Protocol: HTTP
  2. Port: 80
  3. Health check path: / (Not sure if this is correct)
  4. Matcher HTTP Code: 200
  5. Registered target is ECS instance running Graylog container.

Noticed health check status as unhealthy and then change to draining for the target group.

Unfortunately I do not have access to login to the ECS instance to retrieve the docker container logs.


(system) #4

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