[Linux-cluster] Defending GFS

[Ray Charles]

Hi all,

GFS ver 6.1, not gfs2.

I am looking for an intelligent way to defend the use
of gfs on a prod cluster. Every time there's an issue
I hear a cacophony of voices that declare that "GFS
caused ... whatever."

Most recently there was an instance where doing an ls
of a certain sub dir that happened to be on a GFS
partition just hung. All other sub dir in that same
partition could be navigated. Does that sound like a
typical scenario? Aside from the system log is there
any other place to check??  

If gfs gets into a funny state how can i detect it or
test for it?

tia-

-Ray




      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs