<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8">
<META NAME="GENERATOR" CONTENT="GtkHTML/3.18.3">
</HEAD>
<BODY>
On Fri, 2009-03-27 at 17:41 -0600, Gary Sandine wrote:
<BLOCKQUOTE TYPE=CITE>
<PRE>
On Fri, Mar 27, 2009 at 04:58:08PM -0600, Brendan wrote:
> For SCSI, that's true, but SATA (generally) has native hot swap
> capability without notifying the disk subsystem.
Well, I didn't now that. If it's possible to just unplug and replug
without echoing commands into /proc/scsi/scsi, that changes things a
bit. All that's left is mdadm commands, perhaps before and
definitely after the disk change. However, I think it's not as easy
to make a fatal mistake with mdadm as it is with /proc/scsi/scsi.
mdadm seems to catch potential array killing commands and refuse to
do them by
</PRE>
</BLOCKQUOTE>
very interesting, I was totally unaware of that,<BR>
so probably some doc reading previous to the set up and ensuring existence of an spare drive <BR>
(anyhow), (you can always wait until the week end after the spare got used to place another one)<BR>
<BR>
<BLOCKQUOTE TYPE=CITE>
<PRE>
> The only thing worse than having no RAID is, after RAID saves your
> ass, accidentally crashing the system anyhow and being unable to
> restart your RAIDs because a drive is missing ;-)
Yuck.
</PRE>
</BLOCKQUOTE>
<TABLE CELLSPACING="0" CELLPADDING="0" WIDTH="100%">
<TR>
<TD>
<TABLE CELLSPACING="0" CELLPADDING="0" WIDTH="100%">
<TR>
<TD>
-------------------------------<BR>
Andres Paglayan<BR>
<TABLE>
<TR>
<TD>
o:
</TD>
<TD>
(505) 629-4344
</TD>
</TR>
<TR>
<TD>
m:
</TD>
<TD>
(505) 690-2871
</TD>
</TR>
<TR>
<TD>
f:
</TD>
<TD>
(505) 629-1008
</TD>
</TR>
<TR>
<TD>
h:
</TD>
<TD>
(505) 986-1561
</TD>
</TR>
</TABLE>
andres@paglayan.com<BR>
--------------------------------
</TD>
</TR>
<TR>
<TD>
<BR>
<BR>
</TD>
</TR>
</TABLE>
</TD>
</TR>
</TABLE>
</BODY>
</HTML>