1 |
On Tue, 08 Jan 2013 20:06:25 +0100 |
2 |
Florian Philipp <lists@×××××××××××.net> wrote: |
3 |
|
4 |
> Am 08.01.2013 18:35, schrieb Volker Armin Hemmann: |
5 |
> > Am Dienstag, 8. Januar 2013, 08:27:51 schrieb Florian Philipp: |
6 |
> >> Am 08.01.2013 00:20, schrieb Alan McKinnon: |
7 |
> >>> On Mon, 07 Jan 2013 21:11:35 +0100 |
8 |
> >>> |
9 |
> >>> Florian Philipp <lists@×××××××××××.net> wrote: |
10 |
> >>>> Hi list! |
11 |
> >>>> |
12 |
> >>>> I have a use case where I am seriously concerned about bit rot |
13 |
> >>>> [1] and I thought it might be a good idea to start looking for |
14 |
> >>>> it in my own private stuff, too. |
15 |
> >> |
16 |
> >> [...] |
17 |
> >> |
18 |
> >>>> [1] http://en.wikipedia.org/wiki/Bit_rot |
19 |
> [...] |
20 |
> >>> If you mean disk file corruption, then doing it file by file is a |
21 |
> >>> colossal waste of time IMNSHO. You likely have >1,000,000 files. |
22 |
> >>> Are you really going to md5sum each one daily? Really? |
23 |
> >> |
24 |
> >> Well, not daily but often enough that I likely still have a valid |
25 |
> >> copy as a backup. |
26 |
> > |
27 |
> > and who guarantees that the backup is the correct file? |
28 |
> > |
29 |
> |
30 |
> That's why I wanted to store md5sum (or sha2sums). |
31 |
|
32 |
Watch out for circular problems - you will likely store the md5sum on |
33 |
the same medium type you are trying to validate. Which means the md5sum |
34 |
is just as unreliable as the data itself :-) |
35 |
|
36 |
Interesting factoid: we long since passed the point where there is a |
37 |
statistical good chance of cosmic rays flipping bits in a RAID that is |
38 |
being rebuilt *before* the rebuild is complete. Usually we all just |
39 |
pretend it's not like this and we'll get lucky. usually this works out |
40 |
fine because we are lucky |
41 |
|
42 |
-- |
43 |
Alan McKinnon |
44 |
alan.mckinnon@×××××.com |