seena--- via Postfix-users:
> we have a single lmtp end point which is pointing load balancer .
>
> Lmtp endpoint ---> Load balance --> Dovecot director --> Dovecot
> backend
What is an endpoint: TCP socket, UNIX-domain socket, avian carriers,
and how is Postfix configured to CONNECT TO that endpoint? Your
posting has no information about that.
- What is the output from:
postconf -n | grep "lmtp:" #### note the trailing ":"
postconf -M | grep lmtp #### note no trailing ":"
> If any one of the backend failed we are seeing the delay in lmtp
> delivery to even the other backend which are healthy. How to solve this
What does "failed" mean? In other words, What is the error message?
> We need to make sure that only the delivery to the failed backend
> should fail and rest all should work as expected
If a backend becomes unhealthy, then the next connection from Postfix
shoule immediately be routed to a healthy backend, without any delay.
Without further information I cannot distinguish between:
- Postfix is reusing a failed connection. If that is the case,
disabling connection reuse should make a difference.
- The load balancer is not working proiperly.
Wietse
_______________________________________________
Postfix-users mailing list -- postfix-users@postfix.org
To unsubscribe send an email to postfix-users-le...@postfix.org
