Why no RAID-6 in ZimaOS UI?

With 6 HDD drives, why isn’t there an option in the ZimaOS WebUI to create Raid-6, only Raid-0/1/5?

It is no problem creating it via CLI (create an raid-5 with 5 disk in UI, use mdraid to add last drive and grow with level 6 instead of level 5). Most people will use large drives, I have 16TB drives. Raid-5 is just not cutting it, the chance of bit errors while rebuilding a failed drive is to big (relativly)… raid-5 “died” in 2009: Why RAID 5 stops working in 2009 | ZDNET

(the UI still says Raid-5 and the 6th drive has status of “used”:
image

Personalities : [raid6] [raid5] [raid4] 
md0 : active raid6 sdf[6] sde[5] sdd[3] sdc[2] sdb[1] sda[0]
      62502989824 blocks super 1.2 level 6, 512k chunk, algorithm 2 [6/6] [UUUUUU]
      bitmap: 2/117 pages [8KB], 65536KB chunk

unused devices: <none>

root@ZimaCube:/# mdadm --detail /dev/md0
/dev/md0:
           Version : 1.2
     Creation Time : Thu Sep 26 00:09:05 2024
        Raid Level : raid6
        Array Size : 62502989824 (58.21 TiB 64.00 TB)
     Used Dev Size : 15625747456 (14.55 TiB 16.00 TB)
      Raid Devices : 6
     Total Devices : 6
       Persistence : Superblock is persistent

     Intent Bitmap : Internal

       Update Time : Sat Sep 28 00:28:26 2024
             State : clean 
    Active Devices : 6
   Working Devices : 6
    Failed Devices : 0
     Spare Devices : 0

            Layout : left-symmetric
        Chunk Size : 512K

Consistency Policy : bitmap

              Name : zimaos:HDD-RAID
              UUID : b176acff:5f61eb31:4381392e:a2cd7277
            Events : 86695

    Number   Major   Minor   RaidDevice State
       0       8        0        0      active sync   /dev/sda
       1       8       16        1      active sync   /dev/sdb
       2       8       32        2      active sync   /dev/sdc
       3       8       48        3      active sync   /dev/sdd
       5       8       64        4      active sync   /dev/sde
       6       8       80        5      active sync   /dev/sdf

Do we need to send in a feature request or something?

The answer is double parity.

Remember, this is not a commercial product. It’s main purpose (based on ZimaOS) is a home nas, with an assumed primary function of serving up general media.
Double parity means longer write times and loss of storage. Raid 5 is a much more suitable format for a home nas with this assumed primary function. Nobody needs a double parity for 35 seasons of the Simpsons as an example. Raid 5 is a perfect balance of protection and storage for its assumed primary function.

Does the Pro maybe advertise that it can do a bit more than that…sure. But it’s still a consumer product. Perhaps Raid 6 comes along later in the UI, but I think 99% of users are content with the current stock Raid options.

I hear what you are saying, but it is so easy to just add the feature. If you can do raid-5 in the ui, it’s equally simple to do raid-6. it could be under an advanced option for all I care. No need to hide what is available “out of the box = mdraid”. And some people doesn’t want to loose 35 seasons of Simpsons if they can avoid a bit less probability for that when one drive crashed; because one drive will crash, you just don’t know when.
It’s also a Pro version (I have the pro creative one) so why limit me? If it was purely home users, why add NVMe? IceWhale caters more to pro-sumers if you ask me with their thinkerboards (zimaboard/blade), so I don’t buy your reason :slight_smile:
If I want to “loose” 2 drives out of 6 (16TB drives), which I gladly will for peace of mind, and separate backup of important stuff, not Simpsons, thats my choice. Just put up some info telling users what happens with Raid-6.
I’m hoping for raid-6 support in the future, hopefully soon… I now need to consider using manually configured raid-6 or recreate with raid-5 on 5 disks and wait for support in UI for raid-6.

In the mean time, I’ll play with 4x4TB NVMe in slot 7. There I am more happy to use raid-5 for instance, SSD/NVMe has shown to be more reliable, and thus less problem with bit errors when rebuilding.

1 Like

Observation:
If I expand the raid 5 made of 5 of my 6 HDDs to a raid 6 with the last disk, one disk is removed/left out of the raid6 after a reboot. That’s not nice of ZimaOS. I can read and rebuild, but that takes 1-2 days and it’s gone again afgter next reboot.

root@ZimaCube:~# cat /proc/mdstat 
Personalities : [raid0] [raid1] [raid6] [raid5] [raid4] 
md0 : active raid6 sda[0] sde[5] sdd[3] sdc[2] sdb[1]
      62502989824 blocks super 1.2 level 6, 512k chunk, algorithm 2 [6/5] [UUUUU_]
      bitmap: 3/117 pages [12KB], 65536KB chunk

unused devices: <none>

Is there an (unsupported) way to overwrite some config about the raid so this doesn’t happen?