EX4100 Does iSCSI allow use of NTFS formatted drives?


#1

I am getting an EX4100 for use as a Raid storage device. I plan to use bays 1, 2 and 3 as a RAID1 array, and then use bay 4 as a separate JBOD drive for offsite storage.

The one problem with this is that as noted here: https://support.wdc.com/knowledgebase/answer.aspx?ID=1391 if I use “native” mode, the offsite backup drive will be unreadable by any other device since the EX4100 uses a proprietary filesystem, rather than one of the “common” ones, e.g. NTFS, HPFS, ext[234], exFAT etc.

While it’ll be a bit more trouble to set up, can I solve this by using iSCSI for the 4th bay, and then placing an NTFS formatted drive in it and using the built in iSCSI initiator in Windows to access it?

The necessity for this is that data on the device is related to customer accounts, and therefore we may need access to it at a moment’s notice. Waiting a few days for the arrival of a replacement EX device is not acceptable in the event of catastrophic failure.


#2

@dgnuff,

I am not sure the drive uses a proprietary filesystem, but more like the disk is split into multiple partitions and then each partition is formatted with a different partition type. For example on my EX4100, I have these partitions on the first drive:

root@WDMyCloudEX4100 /dev # cat /proc/partitions | grep sda
   8        0 2930266584 sda
   8        1    2097152 sda1
   8        2 2926070784 sda2
   8        3    1049031 sda3
   8        4    1048576 sda4

If I look at the contents of the /proc/mdstat special file:

root@WDMyCloudEX4100 /dev # cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md1 : active raid5 sdc2[0] sda2[3] sdb2[2] sdd2[1]
  8778211776 blocks super 1.0 level 5, 64k chunk, algorithm 2 [4/4] [UUUU]
  bitmap: 2/2 pages [64KB], 131072KB chunk

md0 : active raid1 sda1[3] sdb1[2] sdd1[1] sdc1[0]
  2097088 blocks [4/4] [UUUU]
  bitmap: 0/16 pages [0KB], 8KB chunk

unused devices: <none>

This shows that sda1 and sda2 are used for the RAID devices, where md0 is used for swap (look in the “dmesg” output) and md1 is the actual data RAID drive.

If I do a “df”, I can see that sda4 is used for storing configuration data:

root@WDMyCloudEX4100 /dev # df -k /dev/sda4
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/sda4               951204      1944    932876   0% /mnt/HD_a4

Running the mount command and looking at the “md1” device and the “sda4” device shows the following:

root@WDMyCloudEX4100 /dev # mount | egrep "sda|md1"
/dev/sda4 on /mnt/HD_a4 type ext4 (rw,relatime,data=ordered)
/dev/md1 on /mnt/HD/HD_a2 type ext4 
(rw,noatime,nodiratime,nobarrier,quota,usrquota,grpquota,stripe=48,data=ordered)
/dev/md1 on /mnt/HD/HD_a2/Nas_Prog/_docker/devicemapper type ext4 
(rw,noatime,nodiratime,nobarrier,quota,usrquota,grpquota,stripe=48,data=ordered)

So you can see that “ext4” is used. I am unsure of the contents of /dev/sda3 and if I was more inclined I would probably try and mount it and check it out. From the partition table it is a similar size to partition /dev/sda4.

Not sure if this helps!!!

Cheers,

JediNite


#3

iSCSI would only be a way to present the LUN itself to an iSCSI aware target. How you format that drive/LUN to the target OS is completely up to you. So to answer your question about exclusively using the 4th drive as NTFS, if you can present that entire drive as a LUN to a target via iSCSI, you can format it however you want, and the WD won’t be able to see into the “drive” but will know it’s just an iSCSI device/drive.

For example, I have 4x4TB drives in mine, and have it all as RAID5, so 12TB of useable, and then have a 1 TB LUN, and then 2 400GB LUNs carved out as iSCSI drives, and my WD can’t see into them, but my Windows VMs use them.

Hopefully this makes sense and helps!