1 |
On 03/30/10 18:07, Karl Hiramoto wrote: |
2 |
> On 03/30/2010 04:14 PM, Arkadi Shishlov wrote: |
3 |
>> On 03/30/10 16:28, Karl Hiramoto wrote: |
4 |
>> |
5 |
>>> I've used ext3 and riserfs on CompactFlash but neither seems to be 100% |
6 |
>>> error proof. I've wanted to try NILFS2 but haven't done it yet. |
7 |
>>> |
8 |
>> Let us know then, because something must be done to nilfs2_cleanerd |
9 |
>> which like |
10 |
>> to write a lot. |
11 |
>> |
12 |
> Sounds like a bad idea on flash then. |
13 |
|
14 |
Its probably depends on workload a lot, plus how often cleanerd is allowed to |
15 |
work or is it enabled at all. |
16 |
It actually might be better for flash, cause its not hammering multiple blocks |
17 |
with journal and metadata (sync) or double writes of data=journal. Working in |
18 |
sequential manner is presumably advantageous for (cheap) FTL too. |
19 |
Just a guess... Need someone to try it out in real product. :) |
20 |
|
21 |
The built-in checkpoint/snapshot feature is nice to have. |