1 |
On 30/7/21 10:29 pm, Rich Freeman wrote: |
2 |
> On Fri, Jul 30, 2021 at 1:14 AM William Kenworthy <billk@×××××××××.au> wrote: |
3 |
>> 2. btrfs scrub (a couple of days) |
4 |
>> |
5 |
> Was this a read-only scrub, or did this involve repair (such as after |
6 |
> losing a disk/etc)? |
7 |
> |
8 |
> My understanding of SMR is that it is supposed to perform identically |
9 |
> to CMR for reads. If you've just recently done a bunch of writes I |
10 |
> could see there being some slowdown due to garbage collection (the |
11 |
> drive has a CMR cache which gets written out to the SMR regions), but |
12 |
> other than that I'd think that reads would perform normally. |
13 |
> |
14 |
> Now, writes are a whole different matter and SMR is going to perform |
15 |
> terribly unless it is a host-managed drive (which the consumer drives |
16 |
> aren't), and the filesystem is SMR-aware. I'm not aware of anything |
17 |
> FOSS but in theory a log-based filesystem should do just fine on |
18 |
> host-managed SMR, or at least as well as it would do on CMR (log-based |
19 |
> filesystems tend to get fragmented, which is a problem on non-SSDs |
20 |
> unless the application isn't prone to fragmentation in the first |
21 |
> place, such as for logs). |
22 |
> |
23 |
> Honestly I feel like the whole SMR thing is a missed opportunity, |
24 |
> mainly because manufacturers decided to use it as a way to save a few |
25 |
> bucks instead of as a new technology that can be embraced as long as |
26 |
> you understand its benefits and limitations. One thing I don't get is |
27 |
> why it is showing up on all sorts of smaller drives. I'd think the |
28 |
> main application would be for large drives - maybe a drive that is |
29 |
> 14TB as CMR could have been formatted as 20TB as SMR for the same |
30 |
> price, and somebody could make that trade-off if it was worth it for |
31 |
> the application. Using it on smaller drives where are more likely to |
32 |
> be general-purpose is just going to cause issues for consumers who |
33 |
> have no idea what they're getting into, particularly since the changes |
34 |
> were sneaked into the product line. Somebody really needs to lose |
35 |
> their job over this... |
36 |
> |
37 |
No, it was a normal scrub (read only is an option) - I did the scrub |
38 |
hoping it wasn't necessary but aware that crash halting the OS while |
39 |
doing a backup while the system was generating ooops after an upgrade |
40 |
wasn't going to guarantee a clean shutdown. Ive just kicked off a scrub |
41 |
-r and am getting 41Mb/s speed via the status check (its a usb3 on the |
42 |
disk side, and usb2 on the PC - configuration: driver=usb-storage |
43 |
maxpower=30mA speed=480Mbit/s). I will monitor for a couple of hours and |
44 |
see what happens then swap to a standard scrub and compare the read rate. |
45 |
|
46 |
rattus ~ # date && btrfs scrub status |
47 |
/run/media/wdk/cae17311-19ca-4e3c-b476-304e02c50694 |
48 |
Sat 31 Jul 2021 10:55:43 AWST |
49 |
UUID: cae17311-19ca-4e3c-b476-304e02c50694 |
50 |
Scrub started: Sat Jul 31 10:52:07 2021 |
51 |
Status: running |
52 |
Duration: 0:03:35 |
53 |
Time left: 22:30:40 |
54 |
ETA: Sun Aug 1 09:26:23 2021 |
55 |
Total to scrub: 3.23TiB |
56 |
Bytes scrubbed: 8.75GiB (0.26%) |
57 |
Rate: 41.69MiB/s |
58 |
|
59 |
Error summary: no errors found |
60 |
|
61 |
|
62 |
lsusb: Bus 003 Device 007: ID 0bc2:331a Seagate RSS LLC Desktop HDD 5TB |
63 |
(ST5000DM000) |
64 |
|
65 |
(seagate lists it as a 5Tb drive managed SMR) |
66 |
|
67 |
It was sold as a USB3 4Tb desktop expansion drive, fdisk -l shows "Disk |
68 |
/dev/sde: 3.64 TiB, 4000787029504 bytes, 7814037167 sectors" and Seagate |
69 |
is calling it 5Tb - marketing! |
70 |
|
71 |
BillK |