1 |
Having shot my mouth and hoping to extort some sushi the hardware vs |
2 |
software raid testing should be happening this Sunday. It'll be like the |
3 |
Super Bowl, but for sysadmin geeks. |
4 |
|
5 |
The setup: |
6 |
Dell 2850 |
7 |
2 x Xeon 3GHz w/2MB |
8 |
6 GB RAM in 1GB sticks |
9 |
5 x 73GB 10k U320 |
10 |
PERC4 inboard RAID card w/256 RAM |
11 |
Gentoo 2005.1-r1 gentoo-sources 2.6.15-r1 |
12 |
|
13 |
By pulling the RAM on the PERC you turn it into a normal LSI Megaraid |
14 |
card so the driver and hardware do not change between software and |
15 |
hardware RAID. That should do a nice job of making sure that the way |
16 |
we're doing RAID is the only defference per setup. |
17 |
I have to look at the RAID card and see if it has a battery backup. If |
18 |
not I'll leave write cache on for the disks and the card else I'll only |
19 |
have the card write cache enabled... I may attempt both if there is |
20 |
time. Software RAID will always have write cache on the disks. Towards |
21 |
the end I'll try to do some tests on trying to break write caching on |
22 |
both setups by pulling the power at inopportune times. |
23 |
|
24 |
Cluster size is going to be 64k because that's the current sweet spot |
25 |
and the default on the hardware RAID. If I get time I may also do 32k |
26 |
for comparison. I'd want to try 8k as well in an attempt to throw more |
27 |
data ops at the hardware raid in an attempt to swamp the processor, but |
28 |
only an utter moron would actually use 8k clusters and I definitely |
29 |
won't have time. |
30 |
|
31 |
I may bump the RAM down to 1-2GB so I have less system cache and can |
32 |
keep file sizes down. However I'd prefer to use the full 6 GB so that |
33 |
the 256MB cache on the hardware RAID card is only 5% difference in total |
34 |
RAM vs 25% in a 1 GB system. I may attempt to do both a 1GB and 6GB test |
35 |
if only to see how the extra 256 MB affects things. |
36 |
|
37 |
I'll only be testing RAID 5. That should be a good general case with |
38 |
parity calculations, good read times, and decent write times so we can |
39 |
push each setup fairly hard. I'll be using the same filesystem, ext3, |
40 |
and using the same partitioning scheme in both cases. I will be using a |
41 |
sixth drive as /boot for the software RAID since Linux can't boot RAID 5. |
42 |
|
43 |
I'm still deciding the actual tests. bonnie, iozone, etc are going to |
44 |
be in there, but I'm still looking for some other real world type tests. |
45 |
I'm going to attempt to graph as much system data in each case so we can |
46 |
plot CPU, ram usage, etc as well the actual data from the I/O tests. |
47 |
|
48 |
I think that covers pretty most of what we should care about, but i'd |
49 |
be open to suggestions of things I might have missed. Once testing is |
50 |
done I'll pretty up the graphs, raw data, testing, setup, etc and post |
51 |
it on a web page come Monday or Tuesday. |
52 |
|
53 |
kashani |
54 |
-- |
55 |
gentoo-server@g.o mailing list |