1

So I have two hard drives one connect with SATA and one with USB3.0

SATA (/dev/sdc1):
Apple HDD Toshiba

USB (/dev/sde2):
Western Digital Technologies, Inc. Elements SE Portable

mounted to /mount1 and /mount2 consecutively.


Initial tests

I ran write tests using the command:

dd if=/dev/zero of=testfileR bs=100k count=10000

and read tests using the command:

dd if=testfileR of=/dev/null

/dev/sde2 (/mount1)

write: 1024000000 bytes (1.0 GB) copied, 11.5802 s, 88.4 MB/s
read: 1024000000 bytes (1.0 GB) copied, 1.36778 s, 749 MB/s

/dev/sdc1 (/mount2)

write: 1024000000 bytes (1.0 GB) copied, 11.2735 s, 90.8 MB/s
read: 1024000000 bytes (1.0 GB) copied, 1.32109 s, 775 MB/s


Creating the RAID

Now I am going to create the RAID

mdadm --create /dev/md1 --chunk=256 --level=0 --raid-devices=2 /dev/sdc1 /dev/sde2

Partion:

mkfs.ntfs /dev/md1 -f

RAID test:

/dev/md1 (/RAID2)

write: 1024000000 bytes (1.0 GB) copied, 11.6064 s, 88.2 MB/s
read: 1024000000 bytes (1.0 GB) copied, 1.3568 s, 755 MB/s


Conclusion

Why has the RAID not changed the write speeds?! Am I doing anything wrong?


Strange blkid

/dev/sdc1: UUID="eea8d28a-ad9a-4392-d70a-738c35a9d132" UUID_SUB="dd7d7db3-8c01-9ee6-3808-8d3a3b7dc7d1" LABEL="max-MS-7865:1" TYPE="linux_raid_member" " 
/dev/sde2: UUID="1A6318686F2C1E46" TYPE="ntfs" 

Why is /dev/sde2 not formatted the same as /dev/sdc1 with TYPE="linux_raid_member"?


Output of cat /proc/mdstat

Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10] 
md1 : active raid0 sde2[1] sdc1[0]
      976563968 blocks super 1.2 256k chunks

md0 : active raid0 sdb1[1] sda1[0]
      3907028992 blocks super 1.2 256k chunks

unused devices: <none>
8
  • Reading from /dev/urandom is slower than writing to your disk(s); confirm that dd if=/dev/urandom of=/dev/null still takes as long, then use if=/dev/zero instead. And 80MB probably fits in cache; if so your 'read' isn't actually reading the disk(s). – dave_thompson_085 May 7 '16 at 20:11
  • dd if=/dev/urandom of=/dev/null returns 2.6 MB/s why is it the same? and why so slow? okay I will do the same experiment with /dev/zero – maxisme May 7 '16 at 22:37
  • please see update – maxisme May 7 '16 at 23:12
  • If you're getting read speeds of 750+MB/s from SATA and USB drives, you're not reading from the disk, you're reading from cache. If you want to benchmark drive performance, don't use dd, use something designed for that task, like coker.com.au/bonnie++ – cas May 8 '16 at 2:28
  • please show output of cat /proc/mdstat – cas May 8 '16 at 2:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.