Thank you!

Ed

Because UNAS2 is trash (and its goint to trash)...

Thinking about UNAS4.

Which NIC have UNAS4?

Its also on USB?

Does UNAS4 have that same problems as UNAS2? Carrier problem, SMB freezing, 2.5Gbps drops to 1Gb/FE?

I believe @UI-Team will wont share that important to me informations.

My UNAS4 is with high disk access, and no network usage, for 3 days, since a power outage.

Everything works ok and don't seem to affect performance.

Can't see any process taking too much time (heavier process is smb and rather low) with unify-core appearing sometimes on the top process usage.

@UI-Team Is this normal? Using it for more than 6 months and never happened (but this was the first power outage).

Thanks.

Pedro.

@UI-Team

Please add an optional “Encrypt before upload” setting on backup tasks.

Problem:

Official destinations today copy data as-is:

• Remote UNAS
• SMB/CIFS (including Hetzner Storage Box, Synology, etc.)
• Google Drive / OneDrive / Dropbox
• Amazon S3 / Backblaze B2 / Wasabi

Transport can be encrypted (HTTPS, VPN, SMB seal, rsync daemon/SSH). Destination drive encryption on another UNAS only protects data after it arrives, and only on that UNAS. It also hits the ~143-character path limit.

There is no official way to keep an untrusted offsite target from seeing plaintext files and names. That matters for SMB/S3/B2/Wasabi/Storage Box and any host you do not fully control.

Current workarounds:

1. SSH into the UNAS, configure rclone remotes by hand, wrap the destination in crypt, schedule with cron.
2. Bounce the copy through another NAS that has client-side encryption (Hyper Backup, restic, etc.).

rclone is already on the device and already used by Drive. The first workaround is unsupported: firmware updates have wiped crontab, packages, and configs, and the job never appears in the Backups UI.

Proposed UI:

On Create / Edit Backup Task, for every destination type rclone can speak:

• Toggle: Client-side encryption
• Password + confirm (and optional salt)
• Filename encryption: Off / Obfuscate / Standard
• Directory name encryption: On/Off
• Optional: store the obscured secrets in the existing Drive credential store

Implementation can stay what you already run: a crypt remote wrapping the configured destination. Same rclone sync/copy path as today; only the remote changes from s3:bucket to crypt:s3:bucket.

Restore in the UI would use the same password. Losing the password means the offsite copy is unreadable — that is expected and should be stated clearly, same as drive encryption keys today.

Why this belongs in Drive:

• The binary and the pattern are already there.
• It makes 3-2-1 to cheap object storage / Storage Box actually safe.
• It avoids telling users to SSH in and maintain cron after every UniFi OS / Drive update.
• Encrypted destination volumes on a second UNAS are not a substitute: different threat model, path-length issues, and they do not apply to S3/B2/Wasabi/SMB.

Scope that would already help:

Minimum useful set:

• S3 / B2 / Wasabi
• SMB/CIFS
• SFTP (Storage Box and similar use SFTP on port 23)

Remote UNAS and Google Drive/OneDrive/Dropbox can follow.

Happy to test EA builds. This is the missing checkbox between “Drive backup exists” and “I can put a copy on storage I do not trust.”

@UI-Team

Could someone from Ubiquiti, or anyone with hands-on experience, answer these?

1. Is the ENAS a single controller/mainboard? If it fails, does storage go offline until it's replaced?

2. Is there a supported way to pair two units for HA (active/active, active/passive, or synchronous replication with automatic failover)? If not, is it on the roadmap?

3. What happens to iSCSI LUNs during a firmware update or reboot, and for how long are the hosts' paths interrupted?

4. Do the two 25GbE ports support MPIO to the same LUN, and has that been tested with Hyper-V and Proxmox?

If it's single-controller, that's fine for backup or archive storage. I just need to know whether it's suitable for production VM disks.

Cheers

If I wanted to consolidate my data, plus take back ownership of my data, could I purchase a UNAS device and plug it into my UDM Pro via my 24 Port Pro POE Switch, then remotely access the files when away from my office (I also have Wireguard VPN setup to access my home network when away)?

I would need to find a way to potentially backup the backups offsite so in the event of a major disaster, I wouldn't lose everything. I do have a bank box I have used for storing backups.

We recently migrated a customer from a Synology NAS to a UniFi UNAS using SMB shares. The customer uses Rhino 8 Commercial on Windows.

Rhino can create and save a new .3dm file to the UNAS successfully. However, when saving the same file a second time, Rhino reports:

Failed to save Rhino 8 file. The temporary file could not be renamed.

We verified that this does not appear to be a general SMB permission issue.

From the same Windows workstation and the same SMB folder, the user can successfully:

echo test > rhino-test.tmp
ren rhino-test.tmp rhino-test.txt
del rhino-test.txt

We also verified replacement of an existing file:

echo old > rhino-test.txt
echo new > rhino-test.tmp
move /y rhino-test.tmp rhino-test.txt
type rhino-test.txt

The overwrite/rename succeeds correctly.

Rhino's default advanced setting:

Rhino.Options.FileSettings.WriteLocalTempFileWhenSaving = True

causes the failure.

Changing it to:

Rhino.Options.FileSettings.WriteLocalTempFileWhenSaving = False

completely resolves the issue. Both test files and the original production Rhino files then save normally.

This setting appears to change Rhino's temporary-file/rename save process, so there may be an SMB compatibility difference between UNAS and other NAS implementations during Rhino's atomic/safe-save operation.

Has anyone else seen this with Rhino, CAD applications, or other applications that save using temporary-file replacement?

Environment:

• UniFi UNAS 2
• UniFi Drive / SMB
• Windows 11
• Rhino 8 Commercial
• SMB user permission: Editor

Happy to provide additional logs or testing if useful.

See the previous discussions:

• https://community.ui.com/questions/UNAS2-Slow-Transfer-Speeds/bda74d1f-cebb-49cf-8c5e-74900565e9c4
• https://community.ui.com/questions/UNAS-2-SMB-Issues/94b17ca4-6976-4532-b590-3ccb93c6efe3

In the referenced threads, users reported slow transfer speeds over SMB, as well as hung or stalled transfers. I purchased the UNAS-2-W and began experiencing the same issues.

The root cause appears to be that the UNAS 2's eth0 can spontaneously change from 1 Gb/s to 100 Mb/s or 10 Mb/s.

See the dmesg and ethtool output below.

root@UNAS-2:~# ethtool -S eth0 | grep -Ei 'err|drop|crc|timeout|reset|fail'
   tx_errors: 2
   rx_errors: 1631
   align_errors: 3
   tx_underrun: 0

root@UNAS-2:~# dmesg | grep -Ei 'r8152|eth0|usb|link'
[  0.116416] NET: Registered PF_NETLINK/PF_ROUTE protocol family
[  0.185039] usbcore: registered new interface driver usbfs
[  0.186584] usbcore: registered new interface driver hub
[  0.188093] usbcore: registered new device driver usb
[  0.383213] rtk-usb-manager 98007088.rtk-usb-manager: ENTER rtk_usb_manager_probe
[  0.385358] port_node: usb-port0 status is okay
[  0.386659] rtk-usb-manager 98007088.rtk-usb-manager: port0 mapping gpio_num=560 on gpio0
[  0.388961] port_node: usb-port1 status is okay
[  0.392158] rtk-usb-manager 98007088.rtk-usb-manager: port1 mapping gpio_num=560 on gpio0
[  0.394454] rtk-usb-manager 98007088.rtk-usb-manager: type c connector-switch-gpio no found
[  0.396794] rtk-usb-manager 98007088.rtk-usb-manager: type c plug-side-switch-gpio no found
[  0.399181] port_node: usb-port2 status is okay
[  0.400475] rtk-usb-manager 98007088.rtk-usb-manager: port2 mapping gpio_num=561 on gpio1
[  0.402786] rtk-usb-manager 98007088.rtk-usb-manager: TP1CK_MEM_TEST1_REG=9804ef14 to toggle bit 2 (val=0x501)
[  0.405926] rtk-usb-manager 98007088.rtk-usb-manager: Realtek USB init OK
[  0.407827] rtk-usb-manager 98007088.rtk-usb-manager: __gpio_on_off to set power (low active) off by gpio (id=560) OK
[  0.410786] rtk-usb-manager 98007088.rtk-usb-manager: __gpio_on_off to set power (low active) off by gpio (id=561) OK
[  0.413979] rtk-usb-manager 98007088.rtk-usb-manager: rtk_usb_manager_probe OK (take 32 ms)
[  0.416341] rtk-usb-manager 98007088.rtk-usb-manager: rtk_usb_subnode_probe_work Start ...
[  0.418647] rtk-usb-manager 98007088.rtk-usb-manager: rtk_usb_subnode_probe_work populate subnode device
[  0.426415] rtk-usb-manager 98007088.rtk-usb-manager: Get extcon type_c fail (ret=-517)
[  0.428735] rtk-usb-manager 98007088.rtk-usb-manager: rtk_usb_subnode_probe_work ... End (take 12 ms)
[  0.589026] usbcore: registered new device driver r8152-cfgselector
[  0.590601] usbcore: registered new interface driver r8152
[  0.602112] rtk-dwc3-debug rtk-dwc3-debug.0.auto: dwc3_rtk_debugfs_init for dev=98013200.usb-port0
[  0.604514] rtk-dwc3-debug rtk-dwc3-debug.1.auto: dwc3_rtk_debugfs_init for dev=98013c00.usb-port1
[  0.606874] rtk-dwc3-debug rtk-dwc3-debug.4.auto: dwc3_rtk_debugfs_init for dev=98013e00.usb-port2
[  0.610934] xhci-hcd xhci-hcd.2.auto: new USB bus registered, assigned bus number 1
[  0.612989] xhci-hcd xhci-hcd.2.auto: USB3 root hub has no ports
[  0.618811] hub 1-0:1.0: USB hub found
[  0.620938] rtk-usb-manager 98007088.rtk-usb-manager: gpio_num=560 active_port=0x1 active_port_mask=0x3 ==> No turn on
[  0.625144] xhci-hcd xhci-hcd.3.auto: new USB bus registered, assigned bus number 2
[  0.627182] xhci-hcd xhci-hcd.3.auto: USB3 root hub has no ports
[  0.632984] hub 2-0:1.0: USB hub found
[  0.635078] rtk-usb-manager 98007088.rtk-usb-manager: __gpio_on_off to set power (low active) on by gpio (id=560) OK
[  0.741148] extcon-rtk-type_c 98007220.type-c: First check USB_DR_MODE_PERIPHERAL
[  0.895551] usb 2-1: new full-speed USB device number 2 using xhci-hcd
[  1.455656] rtk-usb-manager 98007088.rtk-usb-manager: rtk_usb_subnode_probe_work Start ...
[  1.481435] rtk-usb-manager 98007088.rtk-usb-manager: __gpio_on_off to set power (low active) off by gpio (id=561) OK
[  1.486539] rtk-usb-manager 98007088.rtk-usb-manager: rtk_usb_subnode_probe_work ... End (take 28 ms)
[  1.569304] xhci-hcd xhci-hcd.5.auto: new USB bus registered, assigned bus number 3
[  1.581077] xhci-hcd xhci-hcd.5.auto: new USB bus registered, assigned bus number 4
[  1.582987] xhci-hcd xhci-hcd.5.auto: Host supports USB 3.0 SuperSpeed
[  1.585148] hub 3-0:1.0: USB hub found
[  1.589324] usb usb4: We don't know the algorithms for LPM for this host, disabling LPM.
[  1.592498] hub 4-0:1.0: USB hub found
[  1.823027] ata1: hard resetting link
[  1.823997] ata2: hard resetting link
[  1.871557] usb 3-1: new high-speed USB device number 2 using xhci-hcd
[  2.064958] hub 3-1:1.0: USB hub found
[  2.138764] ata2: SATA link down (SStatus 0 SControl 300)
[  2.147689] usb 4-1: new SuperSpeed USB device number 2 using xhci-hcd
[  2.192937] hub 4-1:1.0: USB hub found
[  2.483595] usb 4-1.1: new SuperSpeed USB device number 3 using xhci-hcd
[  2.736799] r8152-cfgselector 4-1.1: reset SuperSpeed USB device number 3 using xhci-hcd
[  2.824416] r8152 4-1.1:1.0 eth0: v2.17.2 (2023/09/05)
[  2.825699] r8152 4-1.1:1.0 eth0: This product is covered by one or more of the following patents:
[  7.335545] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[  11.620838] ata2: hard resetting link
[  12.176076] systemd[1]: Listening on Network Service Netlink Socket.
[  13.488831] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[  14.537480] usbcore: registered new interface driver btusb
[  17.072836] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[  17.331551] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[  20.656851] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[  20.661044] r8152 4-1.1:1.0 eth0: carrier on
[  63.129516] usb 2-1: USB disconnect, device number 2
[  64.419570] usb 2-1: new full-speed USB device number 3 using xhci-hcd
[  68.448008] usb 2-1: USB disconnect, device number 3
[  68.811960] usb 2-1: new full-speed USB device number 4 using xhci-hcd
[ 959.426893] r8152 4-1.1:1.0 eth0: carrier off
[ 965.812459] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 969.396472] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 972.984024] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 972.988776] r8152 4-1.1:1.0 eth0: carrier on
[ 9263.585010] r8152 4-1.1:1.0 eth0: carrier off
[ 9269.716104] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9273.300117] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9276.884132] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9276.888321] r8152 4-1.1:1.0 eth0: carrier on
[ 9284.062838] r8152 4-1.1:1.0 eth0: carrier off
[ 9287.380188] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9290.964136] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9294.804200] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9294.809383] r8152 4-1.1:1.0 eth0: carrier on
[ 9333.726412] r8152 4-1.1:1.0 eth0: carrier off
[ 9336.788368] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9340.372374] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9343.959775] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9343.966119] r8152 4-1.1:1.0 eth0: carrier on
[ 9357.022091] r8152 4-1.1:1.0 eth0: carrier off
[ 9360.084449] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9363.668463] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9367.252494] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9367.258627] r8152 4-1.1:1.0 eth0: carrier on
[ 9379.040698] r8152 4-1.1:1.0 eth0: carrier off
[ 9382.100531] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9385.684547] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9389.268576] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9389.273711] r8152 4-1.1:1.0 eth0: carrier on
[ 9687.008358] r8152 4-1.1:1.0 eth0: carrier off
[ 9690.325706] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9693.909720] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9697.493734] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9697.498243] r8152 4-1.1:1.0 eth0: carrier on
[ 9854.945847] r8152 4-1.1:1.0 eth0: carrier off
[ 9858.262345] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[ 9861.846361] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[ 9871.830399] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[ 9871.836329] r8152 4-1.1:1.0 eth0: carrier on
[10707.429597] r8152 4-1.1:1.0 eth0: carrier off
[10713.305606] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[10716.377617] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[10719.961631] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[10719.969860] r8152 4-1.1:1.0 eth0: carrier on
[10956.258540] r8152 4-1.1:1.0 eth0: carrier off
[10967.515422] r8152 4-1.1:1.0 eth0: autoneg retry 1/3 for higher speed (total 1, linked at 100Mbps)
[10979.034617] r8152 4-1.1:1.0 eth0: autoneg retry 2/3 for higher speed (total 2, linked at 100Mbps)
[10988.250654] r8152 4-1.1:1.0 eth0: autoneg retry 3/3 for higher speed (total 3, linked at 1000Mbps)
[10991.834666] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[10991.838281] r8152 4-1.1:1.0 eth0: carrier on
[11078.376065] r8152 4-1.1:1.0 eth0: carrier off
[11089.627095] r8152 4-1.1:1.0 eth0: autoneg retry 1/3 for higher speed (total 1, linked at 100Mbps)
[11092.443054] r8152 4-1.1:1.0 eth0: autoneg retry 2/3 for higher speed (total 2, linked at 1000Mbps)
[11095.515063] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[11095.522256] r8152 4-1.1:1.0 eth0: carrier on
[11191.782797] r8152 4-1.1:1.0 eth0: carrier off
[11195.867445] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[11199.451458] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[11203.035472] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[11203.039727] r8152 4-1.1:1.0 eth0: carrier on
[11209.447553] r8152 4-1.1:1.0 eth0: carrier off
[11220.699531] r8152 4-1.1:1.0 eth0: autoneg retry 1/3 for higher speed (total 1, linked at 100Mbps)
[11229.659571] r8152 4-1.1:1.0 eth0: autoneg retry 2/3 for higher speed (total 2, linked at 1000Mbps)
[11232.731585] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[11232.737084] r8152 4-1.1:1.0 eth0: carrier on
[13271.540620] r8152 4-1.1:1.0 eth0: carrier off
[13292.259433] r8152 4-1.1:1.0 eth0: autoneg retry 1/3 for higher speed (total 1, linked at 10Mbps)
[13310.691506] r8152 4-1.1:1.0 eth0: autoneg retry 2/3 for higher speed (total 2, linked at 10Mbps)
[13329.123573] r8152 4-1.1:1.0 eth0: autoneg retry 3/3 for higher speed (total 3, linked at 10Mbps)
[13347.555643] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 10Mbps
[13347.561550] r8152 4-1.1:1.0 eth0: carrier on
[36799.560009] r8152 4-1.1:1.0 eth0: carrier off
[36815.679649] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[36819.005093] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[36822.333103] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[36822.337999] r8152 4-1.1:1.0 eth0: carrier on
[36907.592791] r8152 4-1.1:1.0 eth0: carrier off
[36910.909435] r8152 4-1.1:1.0 eth0: autoneg retry 1/2 for higher speed (total 1, linked at 1000Mbps)
[36914.493455] r8152 4-1.1:1.0 eth0: autoneg retry 2/2 for higher speed (total 2, linked at 1000Mbps)
[36918.082727] r8152 4-1.1:1.0 eth0: autoneg retry exhausted, accepting 1000Mbps
[36918.091313] r8152 4-1.1:1.0 eth0: carrier on

A few days ago, we got our new UNAS2, which I connected through a UDM SE > USW Pro 24 PoE using the included power adapter.

After getting it set up, I inserted two 3 TB HDDs from an old NAS. I have been trying to add them to the storage pool, but nothing happens. At first, I thought the process might just take a few hours or a day or two, but the drives are still not included in the pool.

There is no progress bar, and after attempting to add them, the system just lets me repeat the exact same process over again which I have tried a few times now without success. The drives themselves are listed as 'optimal'.

Below is what it looks when I add them one by one, followed by the formatting warning, and the last screenshot shows the only information I get. What is wrong here, and why am I not getting any progress indicators or error messages?

With two drive bays, I would like to have the option to choose how the drives are used according to my needs, for example:

• RAID 1, for users who prioritize redundancy;
• RAID 0, for users who prioritize capacity and performance;
• JBOD or independent disks, for users who prefer to use the full combined capacity without mirroring.

In my case, the only thing preventing me from buying the UNAS 2 is this software limitation.

I am not asking for RAID 1 to be removed or changed. On the contrary, it should remain available as an option for users who want disk mirroring and redundancy.

What I would like is simply to have the freedom to choose the storage configuration that best fits my needs.

The hardware already provides two drive bays. Therefore, if this limitation is purely software-related, I hope Ubiquiti will consider adding additional storage configurations in a future UniFi update.

As soon as support for storage configurations beyond RAID 1 becomes available, the UNAS 2 will be an immediate purchase for me.

It would also be helpful to know whether Ubiquiti has any plans to expand the RAID/storage protection options available on the UNAS 2 in the future.

Thank you for considering this request.

Setup:

I don't use the LAG functionality. Each SFP+ interface is dedicated to a discrete Layer 2 VLAN for storage traffic, configured with a static IP and no gateway, since those networks don't need to route anywhere. The copper Ethernet interface is my primary management NIC and is the only interface configured with a gateway.

• enp0s0 (copper, management): 1.2.3.4/24, gateway 1.2.3.1
• enp0s1 (SFP+): 5.6.7.8/24, no gateway
• enp0s2 (SFP+): 9.10.11.12/24, no gateway

Problem:

After some firmware updates or reboots, the UNAS becomes unreachable from outside the storage VLANs. The only way back in is SSH to an SFP IP from a host on the same Layer 2 segment.

Root cause:

UniFi OS installs a default route on every interface, including the SFP interfaces that have no gateway configured. These show up as interface-only default routes (default dev <iface>), all with the same metric (0). When metrics tie, the kernel uses whichever route was installed first, so the active default route depends on interface bring-up order at boot.

Working (copper route first):

default via 1.2.3.1 dev enp0s0 proto static
default dev enp0s2 proto static
default dev enp0s1 proto static

Broken after an update (SFP route first):

default dev enp0s1 proto static
default dev enp0s2 proto static
default via 1.2.3.1 dev enp0s0 proto static

In the broken state, all off-subnet traffic goes out enp0s1 as if every destination were directly attached, so management access is lost.

Workaround:

Manually deleting the gateway-less routes restores access immediately:

ip route del default dev enp0s1
ip route del default dev enp0s2

This doesn't survive reboots, so I'm currently running it from a boot script, which isn't something I should need on a storage appliance.

Requested fix:

Either of these would resolve it:

1. Don't create a default route on an interface that has no gateway configured. This is standard behavior on Linux network stacks.
2. Or add a per-interface route metric (or a "primary interface" setting) in the UI so the management NIC's default route always wins.

Multi-NIC setups with dedicated storage VLANs are a common NAS design, so I'd expect others to hit this too. Happy to provide a support file or test a beta firmware.

Does anybody have problems with UNAS2 like me?

I tried:

Very good RJ45 Cat6a AWG24 cable 0,9m CU

Orginal Ubiquiti RJ45 Premium cable 0,3m

Other patch cable with AWG23 Cat6a ~1m CU

I tried connect directly to USW-Flex-2.5G-8-PoE, and with included Power Injector and then to UCG-Max.

I flash new firmware (also Early Access) and Drive app.

My problems:

• In the past the 2,5Gb link Was going down to 1Gb, and sometimes to 100Mb - but in last month i dont have that problem
• Actually when i check logs in Network i see theres a 2-3-4 times disconnect and connect informations during EACH day. I check logs via ssh, and in last 20h (yesterday i Was manually rebot UNAS2) theres a 48 carrier losts. !!!! 48 TIMES DURING 20 HOURS !!
• Sometimes my SMB share deads. Still I can login via HTTP to UNAS2, its alive, but i have to restart SMB process to able to continue. I made some tweaks via SSH - but ... it doesnt help a lot.

The reason is that theres a cheap Realtek NIC and... its connected to CPU via USB.

Do You understand this? Via USB! My 19" RACK cabinet are cold (~45*C UNAS2 temp during work) ... and it have carrier losts during idle, and freezing SMB process during work.

I just want to take back this... $..T and buy QNAP or Synology.

I really like simple UI and integration with other my Ubiquiti stuff... but... that device...

I cannot re-sell that because Im afraid of that my customer will have problem with them like me.

I read a lot on the internet and theres not only a problem with my device, all of UNAS2 have that problems.

@UI-Team

Can You fix that problems by modyfing firmware, or all UNAS2 from Earth have to go to trash?

Actually i have 2.21.4 driver for RTL8156, can You update to 2.22.x to check it fixes some problems? In relase note theres a note about improve stability and power managment.

1. Does ENAS-to-ENAS replication use native ZFS send/receive (incremental, keeping snapshot history), or file-level copying over SMB like UNAS backups?
2. Can an ENAS replicate to, or receive from, a non-UniFi ZFS system such as TrueNAS?
3. Are cloud backups (S3, B2, Wasabi) file-level or snapshot-based, and is a generic S3-compatible target planned?
4. Is RAIDZ expansion (adding a single drive to an existing group) supported?

Thanks!

UNAS should be doing a reflink copy instead, which is near instant.

After kicking off the large folder move - there is no way to cancel it in the UI. But we were able to cancel it through SSH, let the RAID 1 cache flush, then perform a reflink copy.

Claude Fable understands how the NAS is architected almost instantly. It will be very helpful in wrangling this very immature storage array until Unifi catches up.

Even though the initial storage volume setup is marked as operational and there is currently no active client or file transfer workload, I can hear constant, rhythmic read/write head movement (clicking / seeking sounds) coming from the unit.

The heads appear to be continuously active in the background, creating noticeable drive chatter and vibration through the surface.

1. Does the UNAS 4 run background RAID resynchronization, parity checks, or file indexing upon fresh pool initialization that could cause this persistent head seeking?

2. Is this level of idle head activity expected behavior for high-capacity Toshiba N300 drives in UniFi OS, or should the array settle down once background initialization finishes?

3. Are there diagnostic steps or log views in UniFi OS to verify what specific background task is keeping the heads active?

Any insights from others running the N300 series or fresh UNAS 4 deployments are welcome.

Also, I use the unas 4 in an home environment and there is no way, that it can operate 24h with this level of noise. :(

Can the UNAS 4 be configured as RAID 1 using only two hard drives?

The Drive interface seems to indicate that RAID 1 is supported. It describes one drive being used for data redundancy/mirroring, with the option to add another drive later and upgrade to RAID 5.

I just want to confirm whether a two-drive RAID 1 setup is officially supported on the UNAS 4.

I may upgrade to RAID 5 later, but that is not relevant right now, which is why I am asking specifically about the UNAS 4 rather than the UNAS 2.

Thanks!