1 |
Hi, |
2 |
|
3 |
From the list I already have learned, that most of my concerns regarding |
4 |
the lifetime and maintainance to prolong it are without a |
5 |
reason. |
6 |
|
7 |
Nonetheless I am interested in the technique as such. |
8 |
|
9 |
My SSD (NVme/M2) is ext4 formatted and I found articles on the |
10 |
internet, that it is neither a good idea to activate the "discard" |
11 |
option at mount time nor to do a fstrim either at each file deletion |
12 |
no triggered by a cron job. |
13 |
|
14 |
Since there seems to be a "not so good point in time", when to do a |
15 |
fstrim, I think there must also be a point in time, when it is quite |
16 |
right to fstrim the mu SSD. |
17 |
|
18 |
fstrim clears blocks, which currently are not in use and which |
19 |
contents is != 0. |
20 |
|
21 |
The more unused blocks there are, which has a contents != 0, the |
22 |
lesser the count of blocks is, which the wear leveling algorithm can |
23 |
use for its purpose. |
24 |
|
25 |
That leads to the conclusion: to fstrim as often as possible, to keep the |
26 |
count of empty blocks as high as possible. |
27 |
|
28 |
BUT: Clearing blocks is an action, which includes writes to the cells of |
29 |
the SSD. |
30 |
|
31 |
Which is not that nice. |
32 |
|
33 |
Then, do a fstrim just in the moment, when there is no useable block |
34 |
left. |
35 |
|
36 |
Then the wear-leveling algorithm is already at its limits. |
37 |
|
38 |
Which is not that nice either. |
39 |
|
40 |
The truth - as so often - is somewhere in between. |
41 |
|
42 |
Is it possible to get an information from the SSD, how many blocks are |
43 |
in the state of "has contents" and "is unused" and how many blocks are |
44 |
in the state of "has *no* contents" and "is unused"? |
45 |
|
46 |
Assuming this information is available: Is it possible to find the |
47 |
sweat spot, when to fstrim SSD? |
48 |
|
49 |
Cheers! |
50 |
Meino |