1 |
On Tuesday 24 February 2015 07:31:26 Rich Freeman wrote: |
2 |
|
3 |
> In general though there is a reason that sysadmins tend to be very |
4 |
> conservative with filesystems. I doubt most even jumped onto ext4 all |
5 |
> that quickly even though that was very stable from the start of being |
6 |
> declared as such. You really need to look at your use case and |
7 |
> understand the risks and benefits and how you plan to mitigate the |
8 |
> risks. Something being experimental isn't a reason to automatically |
9 |
> avoid using it if it brings some significant benefit to your design, |
10 |
> as long as you've mitigated the risks. |
11 |
|
12 |
Yes, and that's why I felt the risk justified when I adopted f2fs in |
13 |
that box. It's a LAN server and so doesn't change much, and it's backed |
14 |
up weekly. Well, the web and http-replicator proxies do have constantly |
15 |
changing data of course, but nothing that can't be fetched again |
16 |
cheaply. |
17 |
|
18 |
> And, of course, if your goal is to better understand an experimental |
19 |
> technology in a non-critical setting you should probably just get your |
20 |
> feet wet. |
21 |
|
22 |
Indeed. And I'm sometimes impulsive anyway. I certainly didn't conduct a |
23 |
formal risk assessment. (<tabloid> Shock! Horror! </tabloid>) :-) |
24 |
|
25 |
-- |
26 |
Rgds |
27 |
Peter. |