1 |
On Mon, 2011-03-14 at 23:50 -0700, Matthew Marlowe wrote: |
2 |
> > My problem is that LVM2 is not supported in parted which is the |
3 |
> > recommended tool to deal with this. |
4 |
> > |
5 |
> > I suspect I only need to map the individual PE to a particular start |
6 |
> > sector on each drive, not btrfs, but then there is stripe/block sizes to |
7 |
> > consider as well ... WD also are recommending 1mb sector boundaries for |
8 |
> > best performance - I can see a reinstall coming up :) |
9 |
> > |
10 |
> |
11 |
> I have on my workstation: |
12 |
> 2 WD 2TB Black Drives |
13 |
> 5 WD 2TB RE4 Drives |
14 |
> |
15 |
> Some notes: |
16 |
> - The black drives have horrible reliability, poor sector remapping, and have |
17 |
> certain standard drive features to make them unusable in raid. I would not |
18 |
> buy them again. I'm not sure how similar the green drives are. |
19 |
> - Many of the recent WD drives have a tendency to power down/up frequently |
20 |
> which can reduce drive lifetime (research it and ensure it is set |
21 |
> appropriately for your needs). |
22 |
> - Due to reliability concerns, you'll may need to run smartd to give adequate |
23 |
> pre-failure warnings |
24 |
> |
25 |
> Anyhow, in my config I have: |
26 |
> |
27 |
> 1 RE4 Drive as Server Boot Disk |
28 |
> 4 RE4 Drives in SW RAID10 (extremely good performance and reliability) |
29 |
> 2 Black Drives in LVM RAID0 for disk-to-disk backups (thats about all I trust |
30 |
> them with). |
31 |
> |
32 |
> When I setup the LVM RAID0, I used the following commands to get good |
33 |
> performance: |
34 |
> fdisk (remove all partitions, you don't need them for lvm) |
35 |
> pvcreate --dataalignmentoffset 7s /dev/sdd |
36 |
> pvcreate --dataalignmentoffset 7s /dev/sdf |
37 |
> vgcreate -s 64M -M 2 vgArchive /dev/sdd /dev/sdf |
38 |
> lvcreate -i 2 -l 100%FREE -I 256 -n lvArchive -r auto vgArchive |
39 |
> mkfs.ext4 -c -b 4096 -E stride=64,stripe_width=128 -j -i 1048576 -L |
40 |
> /archive /dev/vgArchive/lvArchive |
41 |
> |
42 |
> I may have the ext4 stride/stripe settings wrong above, I didn't have my |
43 |
> normal notes when I selected them - but the rest of the config I scrounged |
44 |
> from other blogs and seemed to make sense (the --dataalignmentoffset 7s) seems |
45 |
> to be the key. |
46 |
> |
47 |
> My RAID10 drives are configured slightly different w/ 1 partition that starts |
48 |
> on sector 2048 if I remember and extends to the end of the drive. |
49 |
> |
50 |
> The 4 Disk SW RAID10 array gives me 255MB/s reads, 135MB/s block writes, and |
51 |
> 98MB/s rewrites (old test, may need to rerun for latest changes/etc). |
52 |
> |
53 |
> LVM 2 Disk RAID0 gives 303MB/s reads, 190MB/s block writes, and 102MB/s |
54 |
> rewrites (test ran last week). |
55 |
> |
56 |
> Regards, |
57 |
> Matt |
58 |
|
59 |
Thanks Matthew, |
60 |
some good ideas here. I have other partitions on the disks such as |
61 |
swap and rescue so LVM doesnt get all the space. I have steered away |
62 |
from striping as I have lost an occasional disk over the years and worry |
63 |
that a stripe will take out a larger block of data than a linear BOD but |
64 |
your performance numbers look ... great! |
65 |
|
66 |
As the stripe size is hard to change after creation it looks like I'll |
67 |
have to migrate the data and recreate from scratch to get the best out |
68 |
of the hardware. |
69 |
|
70 |
In the short term, I'll just do some shuffling and delete then readd the |
71 |
LVM partition on the green drive to the volume group which should |
72 |
improve the performance a lot. If I am reading it right, I have to get |
73 |
the disk partitioning right first, them make sure the PV is also created |
74 |
at the right boundaries on the LVM. Then I will see how to tune btrfs |
75 |
which I am becoming quite sold on - solid, and online fsck is better |
76 |
than reiserfs which is just as solid, but you have to take offline to |
77 |
check - not that either corrupt often. |
78 |
|
79 |
BillK |