[Info-vax] How would you load balance excess webserver traffic between multiple OpenVMS servers?

[Phillip Helbig undress to reply]

In article <rto3us$evp$1 at dont-email.me>, Dirk Munk <munk at home.nl>
writes: 

> >> let's say you have 4 vms ip interfaces for you server, they can be on 
> >> one server (4 interfaces) or two servers (2 x 2 interfaces) or 4 
> >> servers. The IP addresses are 10.0.0.1 , 10.0.0.2 , 10.0.0.3 , and 
> >> 10.0.0.4 , so very simple.
> >>
> >> With a round robin dns server, you will create a host www.myvms.com , 
> >> and give that host all four IP addresses.
> >>
> >> When you open a connection to www.myvms.com , it will go to 10.0.0.1 . 
> >> A second later it will go to 10.0.0.2 , and again a second later to 
> >> 10.0.0.3 , and then to 10.0.0.4. , and finally back to 10.0.0.1 , and 
> >> so on.
> > 
> > The key question is what happens if 10.0.0.2 is down.
> > 
> > Arne
> > 
> 
> Good point. A good round-robin DNS server would notice that.
> 
> In fact it seems there are round-robin DNS servers that can take the 
> load of a server into account, skipping severely loaded severs for 
> instance. Don't know how that actually works.

On VMS the metric server and the load broker.