<font size=2 face="sans-serif">Kums is our performance guru, so weigh
that appropriately and relative to my own remarks... </font><br><br><font size=2 face="sans-serif">Nevertheless, I still think RAID-5or6
is a poor choice for GPFS metadata. The write cache will NOT mitigate
the read-modify-write problem of a workload that has a random or hop-scotch
access pattern of small writes. In the end you've still got to read
and write several times more disk blocks than you actually set out to modify.
Same goes for any large amount of data that will be written in a
pattern of non-sequential small writes. (Define a small write as
less than a full RAID stripe). </font><br><br><font size=2 face="sans-serif">For sure, non-volatile write caches
are a good thing - but not a be all end all solution.</font><br><br><font size=2 face="sans-serif">Relying on RAID-1 to protect your metadata
may well be easier to administer, but still GPFS replication can be more
robust. </font><br><br><font size=2 face="sans-serif">Doing both - belt and suspenders is
fine -- if you can afford it. Either is buying 2x storage, both is 4x.</font><BR>