[Info-vax] satellite with more than one boot server

[Phillip Helbig---undress to reply]

In article <kejele$co6$1 at dont-email.me>, Stephen Hoffman
<seaohveh at hoffmanlabs.invalid> writes: 

> > A related question: for a satellite with 2 ethernet cards, would it be
> > best to configure 2 identical virtual satellites, differing only in the
> > ethernet address?  They would never be used at the same time.  (I'm
> > assuming that one can't enter more than one ethernet address in the
> > satellite config.)
> 
> What's your goal here?   Complexity?  Uptime?   Screwing around?

Comfort.  If one ethernet card goes flaky, I can just swap the cable and 
reboot.

> In general, my preference would be simplicity.  Leave all but one NIC 
> shut off for MOP activities, probably completely shut off for a low-end 
> case, and configure and activate a secondary NIC only if and as needed. 
>  I might wire it.  My preference here includes having RMC/RCM/MP 
> access, too; remote management access.

It's a DS10, so RMC is there.

> More often than not, adding extra NICs in a cluster can a waste of time 
> and effort and particularly where the benefits of redundancy aren't 
> easily or aren't fully realized, or where you can get tangled in weird 
> partitioned states.  NIC hardware failures are usually system failures. 
>  Switch failures are usually cluster failures or cluster partitions.  
> But the usual NIC failures are not hardware.  They're usually human 
> errors.  Errors such as unplugging or powering down stuff, tripping 
> over network cables, routing errors, etc.  If you do decide to use 
> multple NICs, you'll probably want parallel switches to avoid the 
> switch as a single-point, probably Phase V, and some IP subnet routing.

The idea here is not multiple NICs, separating traffic etc but just to 
have a backup if one fails.