| 1 |
On 2010-05-31, Neil Bothwick <neil@××××××××××.uk> wrote: |
| 2 |
> On Sun, 30 May 2010 14:20:36 +0000 (UTC), Grant Edwards wrote: |
| 3 |
> |
| 4 |
>> > You're assuming that each backup only writes once, which is far from |
| 5 |
>> > true. If you mount a drive with the sync option, the FAT is updated |
| 6 |
>> > for every block you write, so even a single file can cause thousands |
| 7 |
>> > of writes to the same location. |
| 8 |
>> |
| 9 |
>> And you're assuming that the flash controller chip in the USB drive |
| 10 |
>> doesn't do wear-leavelling. |
| 11 |
> |
| 12 |
> Even with wear levelling, writing in sync mode still does thousands of |
| 13 |
> writes. They may be more spread out, but there are still a lot more than |
| 14 |
> one per day and the previous assumptions are still false. |
| 15 |
|
| 16 |
Agreed. Sync writes will definitely wear out flash sooner, but it's |
| 17 |
not as bad as one might think since flash controller chips _generally_ |
| 18 |
do wear levelling and may even do bad-block management that will swap |
| 19 |
in spare blocks when wornw blocks start to go bad. Of couse, none of |
| 20 |
the USB thumb-drive vendors will ever spec any of that, so you have no |
| 21 |
way of actually knowing. |
| 22 |
|
| 23 |
-- |
| 24 |
Grant |
Archives