Disk quantity mismatch between ARM template and scripts

[themorey]

In what area(s)?

/area azlustre

Expected Behavior

I selected 2 disks for the MGS/MDS in the ARM template but the scripts only configured 1 disk...the other is not formatted/used.

Similarly with the OSS I selected 6 disks but only 5 were used.

Actual Behavior

MGS/MDS:

sda     2:0:0:0      30G
├─sda1              500M /boot
├─sda2               29G /
├─sda14               4M
└─sda15             495M /boot/efi
sdb     4:0:0:0       1T
sdc     4:0:0:1       1T /mnt/mgsmds

formatting backing filesystem ldiskfs on /dev/sdc
        target name   LustreFS:MDT0000
        kilobytes     1073741824
        options        -J size=4096 -I 1024 -i 2560 -q -O dirdata,uninit_bg,^extents,dir_nlink,quota,huge_file,flex_bg -E lazy_journal_init -F
mkfs_cmd = mke2fs -j -b 4096 -L LustreFS:MDT0000  -J size=4096 -I 1024 -i 2560 -q -O dirdata,uninit_bg,^extents,dir_nlink,quota,huge_file,flex_bg -E lazy_journal_init -F /dev/sdc 1073741824k

OSS:

sda      2:0:0:0      30G
├─sda1               500M /boot
├─sda2                29G /
├─sda14                4M
└─sda15              495M /boot/efi
sdb      4:0:0:0       1T
sdc      4:0:0:1       1T
└─sdc1              1024G
sdd      4:0:0:2       1T
└─sdd1              1024G
sde      4:0:0:3       1T
└─sde1              1024G
sdf      4:0:0:4       1T
└─sdf1              1024G
sdg      4:0:0:5       1T
└─sdg1              1024G

creating raid (/dev/md10) from 5 devices : /dev/sd[c-m]
devices= /dev/sdc /dev/sdd /dev/sde /dev/sdf /dev/sdg

Steps to Reproduce the Problem

I believe the issue stems from using a VM with no ephemeral disk. In my case I used the Das_v5 so the first available data disk is sdb but that is typically the ephemeral disk.

[themorey]

Created pull request #20 with proposed fix

[edwardsp]

Fixed using Azure symlinks rather than globs from /dev