[mythtv-users] Question re: available SATA ports and linux software RAID

John Drescher drescherjm at gmail.com
Fri Apr 8 15:55:10 UTC 2011

On Fri, Apr 8, 2011 at 11:36 AM, Tyler T <tylernt at gmail.com> wrote:
>> > Just FYI, Linux mdadm software will RAID partitions just as well as disks
>> > (which I have done myself with success on more than one machine). For
>> > example, you could make a small OS partition and large recordings partition
>> > on each of your 2TB drives. Then you can RAID-1 the OS partitions and RAID-6
>> > the recordings partitions.
>> The problem is that you are using the same spindle for your OS (and
>> presumably your DB) and recordings.  That configuration can cause
>> performance issues.
> I agree that with only one spindle this can be an issue. But with 8? ;)

With 8 drives and modern hardware you should have plenty of bandwidth to spare.

Here is a 9 x 2TB hitachi 7200 RPM drive raid6 array which uses the
Intel SASUCI for at least 6 of the 9 ports. The processor is a
core2quad q9550

datastore3 database # uname -a
Linux datastore3 2.6.38-gentoo-r1-mptsas-1 #1 SMP Thu Mar 31 15:50:47
EDT 2011 x86_64 Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz
GenuineIntel GNU/Linux

datastore3 database # dd if=bacula.sql of=/dev/null
16500494+1 records in
16500494+1 records out
8448252980 bytes (8.4 GB) copied, 11.3197 s, 746 MB/s

datastore3 database # dd if=/dev/zero of=data.dat bs=8M count=1000
1000+0 records in
1000+0 records out
8388608000 bytes (8.4 GB) copied, 19.6487 s, 427 MB/s

datastore3 database # dd if=/dev/zero of=data.dat bs=8M count=2000
2000+0 records in
2000+0 records out
16777216000 bytes (17 GB) copied, 41.9033 s, 400 MB/s

datastore3 database # cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4] [raid0] [raid1] [raid10]
md0 : active raid1 sda1[0] sdd1[8](S) sdg1[7] sdj1[6] sde1[5] sdh1[4]
sdf1[3] sdb1[2] sdc1[1]
      272960 blocks [8/8] [UUUUUUUU]

md125 : active raid6 sda5[0] sdh5[6] sdg5[5] sdb5[1] sdj5[8]
sdd5[9](S) sdi5[7] sdf5[4] sde5[3] sdc5[2]
      13617515520 blocks super 1.2 level 6, 128k chunk, algorithm 2

md1 : active raid6 sdg3[5] sdh3[2] sdc3[0] sdf3[7] sdb3[3] sde3[4]
sdd3[9](S) sda3[6] sdi3[1] sdj3[8]
      51447424 blocks level 6, 64k chunk, algorithm 2 [9/9] [UUUUUUUUU]

datastore3 database # pvs
  PV         VG   Fmt  Attr PSize  PFree
  /dev/md125 vg1  lvm2 a-   12.68t 65.27g

datastore3 database # lvs
  LV                   VG   Attr   LSize   Origin Snap%  Move Log Copy%  Convert
  data_analysis        vg1  -wc-ao   1.00t
  database             vg1  -wi-ao 200.00g
  fileserver_imagedata vg1  -wi-ao   3.10t
  imagedata            vg1  -wi-ao   6.00t
  metro                vg1  -wi-ao 100.00g
  portage              vg1  -wi-ao   6.00g
  samba_temp           vg1  -wi-ao   2.00t
  temp_data            vg1  -wi-a-  50.00g
  var_tmp              vg1  -wi-ao  75.00g
  virtualization       vg1  -wi-ao 100.00g

datastore3 database # mount | grep database
/dev/mapper/vg1-database on /mnt/vg1/database type ext4


More information about the mythtv-users mailing list