HAProxy in multi-process mode

Cette page n'est pas encore disponible en français, sa traduction est en cours.
Si vous avez des questions ou des retours sur notre projet de traduction actuel, n'hésitez pas à nous contacter.

When using the HAProxy in multi process mode, each process has its own memory area and therefore has its own statistics.

HAProxy config multi process

This implies that each process needs to have a dedicated socket or endpoint for giving access to its statistics.
Accordingly, in the Datadog configuration file for HAProxy, each socket or endpoint has to be declared as an instance.

HAProxy multiprocess configuration

Otherwise they share the same /haproxy_stats endpoint, and getting the HAProxy statistics from it shows the stats only for the process assigned to the current request.

HAProxy stats 1

Refreshing the page in your browser shows the stats from a different process than previously:

 HAProxy stats 2

If your HAproxy integration is not well configured, you may notice:

  • Missing points on HAProxy metrics that are reported as rate can be checked here and for which you should get a value each 20 seconds.
  • High values and high variations on metrics that are low in normal conditions such as 5xx code error responses.