Script reports to be run from M.2 when using M2. SSD cache on volume

[herbingk]

Script reports to be run from M,2 when using M2. SSD cache on volume. Same occurs for Synology_HDD_DB script as well.

cat /proc/mdstat command returns the following on my DS1621+

Personalities : [raid1] [raid6] [raid5] [raid4] [raidF1] md3 : active raid1 nvme1n1p1[1] nvme0n1p1[0] 1000196800 blocks super 1.2 [2/2] [UU]

md2 : active raid5 sata1p3[0] sata3p3[2] sata4p3[3] sata5p3[4] sata6p3[5] sata2p3[1] 37453707840 blocks super 1.2 level 5, 64k chunk, algorithm 2 [6/6] [UUUUUU]

md1 : active raid1 sata1p2[0] sata4p2[5] sata5p2[4] sata6p2[3] sata3p2[2] sata2p2[1] 2097088 blocks [6/6] [UUUUUU]

md0 : active raid1 sata1p1[0] sata4p1[5] sata5p1[4] sata6p1[3] sata3p1[2] sata2p1[1] 8388544 blocks [6/6] [UUUUUU]

unused devices:

Thanks for looking into it.

[007revad]

Looks like I'll have to setup an NVMe read cache for 1 of my volumes to see what the script is doing wrong.

[herbingk]

Just did a quick check and confirm, by disabling the M.2 SSD cache, the warning disappears.

[007revad]

Ok. I've done some testing and I get the warning if the scriptpath variable is empty or contains a volume# that does not exist.

If you try this test script what does it return? https://github.com/007revad/Synology_enable_Deduplication/blob/test/script_on_ssd.sh

[007revad]

When you run syno_enable_dedupe.sh what does the 4th or 5th line show?

It should show the path and filename of the script, like: Running from: /volume1/scripts/syno_enable_dedupe.sh

[herbingk]

When you run syno_enable_dedupe.sh what does the 4th or 5th line show?

It should show the path and filename of the script, like: Running from: /volume1/scripts/syno_enable_dedupe.sh

Running from: /volume1/admin/Synology_enable_Deduplication/syno_enable_dedupe.sh

[herbingk]

Ok. I've done some testing and I get the warning if the scriptpath variable is empty or contains a volume# that does not exist.

If you try this test script what does it return? https://github.com/007revad/Synology_enable_Deduplication/blob/test/script_on_ssd.sh

/volume1/scripts/test.sh scriptvol: volume1 vg: vg1 md: md3 md2 WARNING Don't store this script on an NVMe volume!

[007revad]

Strange. It doesn't do that for me.

I currently have 4 volumes:

1. volume1 is on HDDs
2. volume3 is on an NVMe drive in the DS1821+
3. volume4 is on a pair of NVMe drives in a E10M20-T1
4. volume5 is on a HDD

And a SATA SSD as a read cache for volume 5.

/volume1/scripts/test.sh
scriptvol: volume1
vg: vg1
md: md3
md2

/volume3/scripts/test.sh
scriptvol: volume3
vg: vg3
md: md4
WARNING Don't store this script on an NVMe volume!

/volume4/scripts/test.sh
scriptvol: volume4
vg: vg4
md: md5
WARNING Don't store this script on an NVMe volume!

/volume5/scripts/test.sh
scriptvol: volume5
vg: vg5
md: md6

I just noticed that you and I are somehow getting an extra md2 on the line after "md: md3" for volume1.

[007revad]

I just ran that test script on my DS720+ which only has 1 HDD volume and I didn't get the warning or the extra md2.

/volume1/scripts/test.sh
scriptvol: volume1
vg: vg1
md: md2

[herbingk]

I just ran that test script on my DS720+ which only has 1 HDD volume and I didn't get the warning or the extra md2.

/volume1/scripts/test.sh
scriptvol: volume1
vg: vg1
md: md2

md3 is my M.2 SSD cache md2 is my RAID5 SATA SSD volume where the script is stored at

cat /proc/mdstat returns:

md3 : active raid1 nvme1n1p1[1] nvme0n1p1[0] 1000196800 blocks super 1.2 [2/2] [UU]

md2 : active raid5 sata1p3[0] sata3p3[2] sata4p3[3] sata5p3[4] sata6p3[5] sata2p3[1] 37453707840 blocks super 1.2 level 5, 64k chunk, algorithm 2 [6/6] [UUUUUU]

md1 : active raid1 sata1p2[0] sata4p2[5] sata5p2[4] sata6p2[3] sata3p2[2] sata2p2[1] 2097088 blocks [6/6] [UUUUUU]

md0 : active raid1 sata1p1[0] sata4p1[5] sata5p1[4] sata6p1[3] sata3p1[2] sata2p1[1] 8388544 blocks [6/6] [UUUUUU]

[007revad]

Your seeing the same as me. md3 is the SSD cache. md2 is the HDD array.

On my DS1821+ I get this:

# lvdisplay | grep /volume_1 | cut -d"/" -f3
vg1

And this:

# pvdisplay | grep -B 1 vg1 | grep /dev/ | cut -d"/" -f3
md3
md2

Digging a little deeper I see this:

# pvdisplay | grep -B 1 vg1
  PV Name               /dev/md3
  VG Name               shared_cache_vg1
--
  PV Name               /dev/md2
  VG Name               vg1

I believe I've found the solution. Can you try this test script to confirm it works correctly for you. https://github.com/007revad/Synology_enable_Deduplication/blob/test/script_on_ssd.sh

[herbingk]

I believe I've found the solution. Can you try this test script to confirm it works correctly for you. https://github.com/007revad/Synology_enable_Deduplication/blob/test/script_on_ssd.sh

Sure, returns no more warning:

/volume1/scripts/test.sh scriptvol: volume1 vg: vg1 md: md2

[007revad]

Fixed in https://github.com/007revad/Synology_enable_Deduplication/releases/tag/v1.2.17

Thank you.

[herbingk]

It was a pleasure to me. I have to thank you for creating, maintaining and sharing these valuable scripts with us. Very much appreciated!