[Info-vax] Distant Cluster?
Christoph Gartmann
gartmann at nonsense.immunbio.mpg.de
Mon Oct 8 05:54:50 EDT 2012
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.
>> Now it is possible to shutdown all the Alphas, the Itanium boxes continue to
>> run and vice versa. But if the network connection between the two sites drops,
>> the Itaniums crash. Why?
>
>Because you haven't yet configured a stable network, or haven't added a
>parallel/redundant connection? Or because your network is getting
>overloaded, and the load of (for instance) shadowing is saturating the
>wire? Or wasn't that the (intended) question?
There are a few switches between the two locations. These need software updates
from time to time and require a reboot...
>FWIW, the hosts usually crash when the network connection resumes,
>because they're determined to the "outliers" in the cluster and
>CLUEXIT; you ended up with a partition, and the Alpha boxes "won" the
>decision during the reconnection process.
>
>Why does this happen? Partitioning. Or more specifically, avoiding
>partitioning.
Ok, I understand.
>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.
Regards,
Christoph Gartmann
--
Max-Planck-Institut fuer Phone : +49-761-5108-464 Fax: -80464
Immunbiologie und Epigenetik
Postfach 1169 Internet: gartmann at immunbio dot mpg dot de
D-79011 Freiburg, Germany
http://www.immunbio.mpg.de/home/menue.html
More information about the Info-vax
mailing list