Mirror SSD as Proxmox installation storage issue

[ieronymous]

Hello

I ve recently noticed 3 things with my 2 SSDs running in mirror created during prox installation.

First I noticed that in the GUI -> Node->Disks, under the tab usage each ssd has the <<partitions>> word (installation made with mbr method)
In another proxmox installation again with 2 ssds in mirror mode, it shows under GUI -> Node->Disks zfs_member under Usage tab (installation made with uefi mode)
Why is that naming scheme difference?

Second thing I ve noticed is that I have in both (different) nodes a message while running zpool status displaying
<<
status: Some supported features are not enabled on the pool. The pool can
still be used, but some features are unavailable.
action: Enable all features using 'zpool upgrade'. Once this is done,
the pool may no longer be accessible by software that does not support
the features. See zpool-features(5) for details.
>>
Does anyone else having this message and did he run the zpool upgrade command?

Last but not least, even though the wearout level is 0% (reminding you that they are both new drives with less than a day hours) for both of them (same WD probably enterprise drives 128gb each model HBS3A1912A7E6B1) if i press in each of them the show SMART values I get a line Media Wearout indicator with a value of 10 (before a while it as 9)

Does anyone have any experience with the above issues?

Thank you

PS Maybe important but forgot to mention that since the workstation DellT7610 has no official room for 10 drives (which I managed to fit) I have one ssd connected to the motherboard via a pci-e card from OWC and the other one via the odd (by removing the cd rom and changing to an hdd caddy)

 

[ph0x]

First I noticed that in the GUI -> Node->Disks, under the tab usage each ssd has the <<partitions>> word (installation made with mbr method)
In another proxmox installation again with 2 ssds in mirror mode, it shows under GUI -> Node->Disks zfs_member under Usage tab (installation made with uefi mode)
Why is that naming scheme difference?

With my installations it's "partitions" whenever Proxmox was installed on the disks (because of other partitions lying on them, like efi) and "zfs member" when it's a whole disk for zfs. Are you sure that these are system disks that say "zfs member"?

Second thing I ve noticed is that I have in both (different) nodes a message while running zpool status displaying
<<
status: Some supported features are not enabled on the pool. The pool can
still be used, but some features are unavailable.
action: Enable all features using 'zpool upgrade'. Once this is done,
the pool may no longer be accessible by software that does not support
the features. See zpool-features(5) for details.
>>
Does anyone else having this message and did he run the zpool upgrade command?

I ran it on all of my pools and didn't have any problems. It comes from the newest version of zfs.

Last but not least, even though the wearout level is 0% (reminding you that they are both new drives with less than a day hours) for both of them (same WD probably enterprise drives 128gb each model HBS3A1912A7E6B1) if i press in each of them the show SMART values I get a line Media Wearout indicator with a value of 10 (before a while it as 9)

Don't look at the raw values, they change a lot and are often manufacturer specific. It's the second column of numbers which is relevant and that shows 100 (%), so nothing to worry about.

 

[ieronymous]

With my installations it's "partitions" whenever Proxmox was installed on the disks (because of other partitions lying on them, like efi) and "zfs member" when it's a whole disk for zfs. Are you sure that these are system disks that say "zfs member"?

Pretty sure as you can see in the attached picture zfs_member.jpg

I just re installed prox using uefi and (probably should have mentioned it earlier in case it matters) during installation as in the previous install attempt, I used advanced and se 100gb from the 128gb for over-provisioning
Another weird thing is that even though during installation the boot drives were sde and sdf somehow in the gui states the one as sda and the other still remains sdf (check photo disk_name_type.jpg ). The type of these drives also changed to BIOS BOOT Why? (Not usign grub - black and white screen during boot and the blue one)

I ran it on all of my pools and didn't have any problems. It comes from the newest version of zfs.

I tried that too and after issuing command
zpool upgrade rpool
I am getting as a reply
This system supports ZFS pool feature flags.

cannot set property for 'rpool': invalid feature 'redaction_bookmarks'
and stops there...

Edit: In the second node with less quality ssd drives zpool upgrade rpool gives me
This system supports ZFS pool feature flags.
Pool 'rpool' already has all supported features enabled.

NEW EDIT!! After a reboot and trying again
:~# zpool upgrade rpool
This system supports ZFS pool feature flags.

Enabled the following features on 'rpool':
redaction_bookmarks
' redacted_datasets
bookmark_written
log_spacemap
livelist
device_rebuild
zstd_compress

:~# zpool version
zfs-2.0.4-pve1 (remained the same)
zfs-kmod-2.0.4-pve1 (it changed from 0.8.5 to 2.0.4)

Don't look at the raw values, they change a lot and are often manufacturer specific. It's the second column of numbers which is relevant and that shows 100 (%), so nothing to worry about.

I thought that this was relevant to Seagate drives only and not others. By the way this media wear indicator (which as an option isnt available to my other plain ssd drives went from 4->12 and 9->18 I wonder what the heck it is calculating)