We have a NodeJs application (Expressjs framework) in two environments staging and production. We are using Winston library for logging to Logstash.
We have two opened ports on Logstash, one for production and another for staging.
-
Logstash server and the production server are connected through a private network. so the connection between them is using private IPs.
-
Logstash server and the staging server are on a public network. so the connection between them is using public IPs.
-
in both environments, we are using the UDP protocol to send logs to Logstash.
The problem:
There is a stringent message size limit on logs from staging to Logstash, and the limit is very small. However, this limit doesn't exist between production and Logstash, although we are using the same config for the two ports on Logstash and we are using the same code for logging using Winston.
So the case now is almost all staging logs are not received by the Logstash. on the contrary production, logs are received as expected.
Questions:
- Why would there be a tiny limit on the message size? is there a way to override that? why it is only happening only in the staging env?
- What are the aspects that affect the message to be received by the Logstash?