to be correct,
slots 3 and 4 (not 2 and 3) must be populated with disks
you count 3 and 4 slots starting from 0, from the right to the left

also, it is advise to remove a broken edisk from the shelf as it could parasite the fiber channel loop
that's all the more since it is the cotroller board which is dead in the disk

try to put move a spare disk to the slot of the broken removed disk if this disk was in slot 3 or 4 as soon as possible


Daniel Finn wrote:
Interesting.  Not slots are empty on these.

-----Original Message-----
From: Dirk Schmiedt [mailto:Dirk.Schmiedt@munich.netsurf.de]
Sent: Tuesday, April 29, 2003 2:47 PM
To: Daniel Finn
Cc: 'toasters@mathworks.com'
Subject: Re: i need help making sense of what happened here


Daniel Finn wrote:

  
Today a disk failed, but different from what I'm used to seeing.  It
    
appears
  
now that I look at it that disk 4.19 failed and it took spare disk 4.38 and
began rebuilding on that.  What seems strange though is that it kept trying
to read or write to disk 4.19 until I physically pulled the disk out.  Why
would this have happened?

    
Hello Daniel

It appears, that the electronic controller of your disk failed.
The filer tried to tell the disk "you are broken" and tried to write on 
the disks label, to make shure wherever this disk will go to ... it will 
tell any technician by it's raidlabel: I am broken.
In your case I also see SES error messages for this shelf.   ... let's 
calculate: Disk 19=16+3
Let me guess ... This is a FC-7/8/9 shelf and the slot 2 is/was empty?  
The slots 2 AND 3 should be filled both in those old schelves, just in 
case one of those disks breakes...

I think you can ignore the warnings you have seen. They were produced by 
the disks electronic.

Smile & regards
Dirk



  

-- 
-- jetez un oeil ici http://www.sensiva.com/ --
--  have a look here http://www.sensiva.com/ --