1 |
On Monday 01 February 2010 14:20:28 Stroller wrote: |
2 |
> On 1 Feb 2010, at 11:58, J. Roeleveld wrote: |
3 |
> > ... |
4 |
> > I am currently installing a new server and am using Linux software |
5 |
> > raid to |
6 |
> > merge 6 * 1.5TB drives in a RAID5 configuration. |
7 |
> > |
8 |
> > Creating the RAID5 takes over 20 hours (according to " cat /proc/ |
9 |
> > mdstat ") |
10 |
> > |
11 |
> > Is there a way that will speed this up? The drives are new, but |
12 |
> > contain random |
13 |
> > data left over from some speed and reliability tests I did. I don't |
14 |
> > care about |
15 |
> > keeping the current 'data', as long as when the array is reliable |
16 |
> > later. |
17 |
> > |
18 |
> > Can I use the " --assume-clean " option with mdadm and then expect |
19 |
> > it to keep |
20 |
> > working, even through reboots? |
21 |
> > Or is this a really bad idea? |
22 |
> |
23 |
> It wasn't my intention to chide you - I don't use software RAID |
24 |
> myself, and your question piqued my curiosity - but the first three |
25 |
> Google hits for "assume-clean" indicate that this isn't safe to use |
26 |
> with RAID5. |
27 |
> |
28 |
> The 4th Google hit contains an extract from the manpage: |
29 |
> |
30 |
> ... It can |
31 |
> also be used when creating a RAID1 or RAID10 if you want |
32 |
> to avoid the initial resync, however this practice -- |
33 |
> while normally safe -- is not recommended. Use this |
34 |
> only if you really know what you are doing. |
35 |
|
36 |
I did find the same results on Google, but not really a proper explanation as |
37 |
to why it's a "bad idea". Unfortunately, my budget doesn't extend to a |
38 |
hardware raid solution. (The cheap cards offload it to the CPU anyway and are |
39 |
generally considered slower in various benchmarks) |
40 |
|
41 |
> I kinda expected this 20 hours to be spent verifying that the disks |
42 |
> contain no bad sectors, which would really hose you if it were the case. |
43 |
|
44 |
True, but I already ran "badblocks" twice on each disk to verify that the |
45 |
disks are fine. (No badblocks found). |
46 |
|
47 |
> But OTOH, 20 hours does not seem an outrageous amount of time for |
48 |
> building a 7.5TB array. You're not going to do this often, and you |
49 |
> want it done right. |
50 |
|
51 |
Good point, and I agree, which is why I will let it finish it's course, but I |
52 |
also expected it could be done quicker. |
53 |
|
54 |
> It would be interesting to know whether hardware RAID would behave any |
55 |
> differently or allow the sync to perform in the background. I have |
56 |
> only 1.5TB in RAID5 across 4 x 500gb drives at present; IIRC the |
57 |
> expansion from 3 x drives took some hours, but I can't recall the |
58 |
> initial setup. |
59 |
|
60 |
I'm hoping someone with more knowledge about RAID-systems can throw in his/her |
61 |
2cents. |
62 |
|
63 |
Thanks, |
64 |
|
65 |
Joost |