[Info-vax] Distant Cluster?

[Stephen Hoffman]

On 2012-10-08 09:54:50 +0000, Christoph Gartmann said:

> In article <k4pl2v$kks$1 at dont-email.me>, Stephen Hoffman 
> <seaohveh at hoffmanlabs.invalid> writes:
> 
>> You mention "distant" in the title.  How distant?  The
>> officially-supported cluster distance is 500 miles / 800 kilometers
>> with the HP default cluster configuration support.  Longer spans can
>> and do work, though you'll (officially) want to work with HP if you
>> need formal support.
> 
> Approximately 500 meters, still within the same LAN.

This is more a case of measured latency and bandwidth than distance.  
The building-to-building distance solely provides the theoretical 
minimum latency and theoretical maximum bandwidth.

> ...
> There are a few switches between the two locations. These need software updates
> from time to time and require a reboot...
> ...
> A switch requires 2 minutes to reboot. So I think this is a bit too long.
> I will think about alternate network paths.


Either multiple network paths, or switches that recover more quickly, 
or switches that don't need firmware updates quite as often and/or 
updates with some coordination of the switch outages so that you can 
shut down the lower-voting lobe.  Or crank up the cluster reconnection 
interval to two minutes and change, and hope to ride over.  And FWIW 
multiple network paths can trigger outages due to the length of time 
that the spanning tree algorithm takes to resolve itself, depending on 
how capable the switches are, and how the switches are configured.


-- 
Pure Personal Opinion | HoffmanLabs LLC