I agree that there may be no "right" answer, but there are an awful lot of 
wrong ones.  


Please don't take this as a criticism of the way you are set up, I'm sure it 
works fine for you.  The problem I see with keeping everything on a single 
volume is that you are creating a single point of failure.  Because of the 
sheer size of keeping everything on one pack you are extending both recovery 
and the necessary backups.  The zfs datasets are too volatile (my opinion only) 
to be kept there.  I know of sites that believe that they should keep their 
RES, DLB and USS stuff on the one volume.  If it works for them, then I'm happy 
for them, but I would not do that.  

If it works for you, then I am very happy for you and wish you luck, but I 
would not be able to sleep well knowing that someone on the system programming 
staff can make a brain check error and ruin everyone's day.  Or that an 
application programmer messes up and one of my zfs's is now 900,000,000 
cylinders. :)

I really should not comment like this, because I truly hate when people don't 
agree with me, but I didn't want to let the OP think that this was was 
acceptable, or simple, from a backup and recovery basis.  

Please don't take my comments badly.  I'm just trying to give the OP a rule of 
thumb that will fit any site and be easy to maintain.

Brian

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to