1 |
On Fri, 24 Feb 2012 20:55:21 +0000 |
2 |
James Broadhead <jamesbroadhead@×××××.com> wrote: |
3 |
|
4 |
> On 23 February 2012 21:29, Grant <emailgrant@×××××.com> wrote: |
5 |
> > [snip] |
6 |
> > |
7 |
> > I'm amazed but disconnecting and reconnecting the IDE and power |
8 |
> > cable fixed it. Which is your favorite tool for testing a HD's |
9 |
> > integrity with and without S.M.A.R.T. support? |
10 |
> |
11 |
> [I] gnome-extra/gsmartcontrol [1] |
12 |
> Available versions: (~)0.8.6 {debug} |
13 |
> Installed versions: 0.8.6(16:47:27 13/02/12)(-debug) |
14 |
> Homepage: http://gsmartcontrol.berlios.de/ |
15 |
> Description: Graphical user interface for smartctl |
16 |
> |
17 |
> [1] "sunrise" /var/lib/layman/sunrise |
18 |
> |
19 |
> Is a great (and sorely needed) frontend for smartmontools - it even |
20 |
> colours lines in red when they indicate imminent failure! |
21 |
> |
22 |
> Make sure that you have read the Google paper before trusting SMART |
23 |
> too far though -- they found (among other things) that it only |
24 |
> accurately predicts failure in 50% of cases. |
25 |
> |
26 |
|
27 |
That's a fair trade in my book, seeing as without it I could |
28 |
manage to accurately predict failure in 0% of cases. |
29 |
|
30 |
Now to get the managers to understand that it's 50%, not 100% and stop |
31 |
the endless whinging when SAN drives fail without prior SMART alerts |
32 |
|
33 |
While on the topic, has anyone heard of research into false positives |
34 |
wrt S.M.A.R.T? |
35 |
|
36 |
-- |
37 |
Alan McKinnnon |
38 |
alan.mckinnon@×××××.com |