Fas920c Shelf count mismatch

[ploughingon]

hi ,
we are getting a shelf count mismatch confguration error from our cluster.
one filer has normal global status
the warns us that it is unable to takeover the first

all services hanging of them is fine so they both working.

how can i tell from the command line if a filer is panicing? I'm guessing it is and its going to shutdown but i could really do with knwoing how long i have befoer it does this..

thanks

 

[xmsre]

A shelf count mismatch that does not resolve within a minute or so usually indicates the failure of a shelf, cable, or adapter. You need to investigate further to determine which component has failed.

 

[RMGBELGIUM]

run "storage show disk -p" This should show 2 paths per disk.If somewhere it doesn't, then , following things could be broken :
- port of HBA
- fiber from HBA to first shelf in loop ( if a whole diskloop doesn't have 2 paths )
- copper cable ( if 1 or more shelves show disks with only 1 path in a loop)
- ESH module

Or if it is a 1 time occuring error, someone could have pulled and replugged a fiber,so then it is nothing to worry about

rgds,

R.