> Is esb2:38181 the instance you bring down or second
> instance on which
> request need to be failed over? If it is second
> instance, then until it
> is detected to be healthy, you will see blue screen
> as there is no
> healthy instance to service your request. In such a
> case, I don't see
> any logs for instance1 which is brought down by you.
>
> I am not sure why you are not able to telnet. May be
> some system/network
> configuration issue?
Thank you for the reply. The problem is that I do not do anything neither with the glassfish cluster nor with the webserver. Since I wrote my first post, behaviour described in it has not changed: lbhealthchecker keeps on detecting daemon
https://esb2:38181 as unhealthy, and after a few seconds it is detected as healthy. I presume that the reason for this could be glassfish https daemon on machine esb2 being somehow brought down and after a while up again. This would explain why I cannot telnet onto port 38181 on machine esb2 and after a few seconds I connect without a problem. The other option is, as you mentioned, the problem with network, but as far as I remember I got the same problem on similar configuration on virtualbox machines. Moreover, if it was a problem with network, the daemon listening on port 38080 on machine esb2 would be brought down as well, wouldn't he? Maybe there is something wrong with ssl? If so, it is probably worth mentioning that i did not manage do configure automatic loadbalancing.xml export to webserver through network. Instead, I copy that file manually.
[Message sent by forum member 'timtowdi']
http://forums.java.net/jive/thread.jspa?messageID=395450