[Info-vax] Distant Cluster?

[Michael Moroney]

gartmann at nonsense.immunbio.mpg.de (Christoph Gartmann) writes:

>There are a few switches between the two locations. These need software updates
>from time to time and require a reboot...

>>The ugly approach is to increase the cluster timeout past your 
>>networking group's typical network outage interval; see RECNXINTERVAL 
>>here 
>><http://h71000.www7.hp.com/doc/84final/4477/4477pro_026.html#clus_sysgen> 

>A switch requires 2 minutes to reboot. So I think this is a bit too long.
>I will think about alternate network paths.

Many systems manager try to get a "private" link between cluster members
for the cluster traffic with no, or as few as possible, switches between
them, and use the "public" (organization-wide) LAN for backup only or not
at all for cluster traffic. (of course it is used for non-cluster traffic,
e.g. TCP/IP)

What speed do you use? 100base-FX full duplex and fibre versions of 
gigabit can be done as a single run within your 500m distance.

Of course this may be difficult or impossible in your organization.

A hack alternative, if you know in advance a switch will be rebooted,
is to temporarily set RECNXINTERVAL to longer than the switch reboot
time (on all nodes) and set it back to a more sane value when done.
RECNXINTERVAL is dynamic.