SATA link-reset errors - resolved

["Rich Braun" <user=44@nabble.blu.org>]

 A couple days ago I posted a bit about my new server and now that I've 
resolved one of my obscure issues, I wanted to help out anyone else who runs 
into this problem on a low-power server. 

Basically what I wanted was to have 7 drives installed and configure the 
arrays in such a way as to split spindles among applications so that, 
effectively, when I'm not watching TV the most-possible drives will be spun 
down to save electricity. 

I noticed two problems:  (1) the drives weren't staying spun down and (2) 
there were lots of link-reset errors in syslog. 

On closer examination, I found that the errors (aside from one truly-failing 
drive) were coming out once every half hour. 

Cutting to the chase:  if you want to activate a spin-down timer (hdparm -S 
xxx) on your drives, at least in the case of the stock openSUSE 11.0 distro, 
you also have to add the "-n standby" option on the DEVICESCAN line of your 
/etc/smartd.conf. 

-rich 
P.S.  I'm still hiring for the team at manhunt.net.  I've gotten two very good 
Linux gurus already in recent months and it's time to add two more (a MySQL db 
guru and another systems engineer).  Write me if you're thinking of a career 
jump; we're really doing some very different things with Linux systems admin. 


-- 
This message has been scanned for viruses and 
dangerous content by MailScanner, and is 
believed to be clean. 

_______________________________________________ 
Discuss mailing list 
[hidden email] 
http://lists.blu.org/mailman/listinfo/discuss