qpid-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gary Tully <gary.tu...@gmail.com>
Subject Re: No action reqiured: reconnect/retry observation
Date Mon, 03 Jul 2017 12:05:00 GMT
that has been the experience in activemq. see:
http://activemq.apache.org/failover-transport-reference.html

the number of attempts and the delay between retry attempts can be
different for the initial connection and "re" connection.

On Fri, 30 Jun 2017 at 22:12 Alan Conway <aconway@redhat.com> wrote:

> This is for when we get to defining a general reconnect strategy, not
> for immediate action. Satellite's reconnect issue has made me realise
> this:
>
> Need to distinguish "connection retry" and "reconnect", may need
> different behaviour:
>
> 1. *retry*: initial connection attempt fails, try again.
> 2. *reconnect*: established connection breaks, re-establish.
>
> Important differences:
>
> 1. may involve security, misconfiguration or other permanent errors
> that require application intervention.
> 2. is a network failure or server fail-over, more likely to be
> recoverable by automatic re-connect.
>
> 1. no state to be recover on retry, since there was never a connection.
> 2. may need to re-establish sessions/links, re-send messages, and
> let application rebuild state.
>
> Cheers,
> Alan.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@qpid.apache.org
> For additional commands, e-mail: users-help@qpid.apache.org
>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message