users@glassfish.java.net

Re: Restarting a cluster seems to break load balancer, is this a bug?

From: Kshitiz Saxena <Kshitiz.Saxena_at_Sun.COM>
Date: Tue, 11 Dec 2007 23:13:41 +0530

Hi,

Have you configured health-checker? If health-checker is not configured,
load-balancer can never detect instance health once it goes down.

Please add below entry to each cluster :
<health-checker url="/" interval-in-seconds="120" timeout-in-seconds="30" />

Thanks,
Kshitiz

glassfish_at_javadesktop.org wrote:
> Hi,
> I am using SJSAS 9.1. There are two clusters that are running fine. I configured the Http Load Balancer to forward requests to both of these clusters. Every thing works fine until I restart one of the clusters.
> When I stop and start a cluster, the load balancer doesn't forward the requests anymore to that cluster. I get the "Due to a temporary error the request could not be serviced" blue screen. Is this a bug or am I missing something obvious. The applications on the cluster are all up and running.
> If I reload the loadbalancer.xml, then it seem to work fine again.
>
> Thanks for helping
> [Message sent by forum member 'kingdavid' (kingdavid)]
>
> http://forums.java.net/jive/thread.jspa?messageID=249531
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: users-help_at_glassfish.dev.java.net
>
>