I share my Drives over SMB, and have enabled the SMB Trash option. This works fine for keeping deleted files in the Trash.

But the Restore option is only available at the drive root level. If a file is deleted in any other path, the only option is to Download it.

Can you please add Restore to all files in any path?

When Restored, I would expect the path to be recreated as required and the file to be restored to the original location.

Thank you!

just bought my UNAS Pro 4 and it works great. But I was shocked, when I found out, that an SMB Offsite Backup ist not encrypted and that Unifi Drive 4 has no encryption für Backups whatssoever.

I highly request that feature. It is a shame that Ubiquity releases "Enterprise" Hardware and you are not able to encrypt you Backups.

If, when will this feature will be available ?

Greetings from Germany

Wally

Fortunately, standard rsync daemon mode (TCP port 873) works perfectly, and the necessary rsync service and configuration already exist on the UNAS Pro. All I needed to do via SSH was:

systemctl start rsync
systemctl enable rsync

My existing ReadyNAS backup jobs immediately started working again, using the rsync accounts and modules already configured through UniFi Drive.

The problem: Firmware updates can disable the rsync service again, breaking scheduled backups with Connection refused (111). Enabling the service survives normal reboots but isn't reliably preserved across firmware updates.

For context, this is authenticated NAS-to-NAS backup traffic on a trusted local network. TCP port 873 can be restricted to specific source IPs. I'm not requesting Internet-facing rsync access or suggesting daemon mode as a replacement for encrypted SSH transfers.

@UI-Team : Could you please officially support rsync daemon mode (TCP 873), ideally with an enable/disable toggle under UniFi Drive → Settings → Services → Rsync?

It would be reasonable to leave daemon mode disabled by default, while allowing administrators to explicitly enable it, configure authentication and source-IP restrictions, and have that setting persist through firmware upgrades.

The functionality is already there and works. It just needs official support and a persistent configuration option.

how does backup work?

is it onetime or will it periodically update based on added, removed, or updated files?

I am looking for a way to do an initial backup, and then periodically update what has changed daily.

I need to have multiple complete backups.

Thank you

Custom alarms simply arent triggering. Attached are screenshots showing the test alarm I built, the built in alarms triggering and not the custom one, and the log shared drives are being created.

Is this a bug?

Im finding the fan control a bit poor? with the settings at Balanced the drives are reaching almost 48degrees at idle with nothing happening.

If i set t the fan control to "cooling" im getting around 23 degrees but the thing sounds like a rocket taking off.

Is there no other planned control? or able to adjust it manually? seems disapointing.

This worked for me and mapped the right user/group.

sudo mount -t cifs //<UNAS_IP>/<SHARE> <MOUNTPOINT> -o credentials=/etc/samba/unas-credentials.txt,uid=1000,gid=1000,forceuid,forcegid

Related: UNVR G2 (Protect 7.3.70, UniFi OS 6.0.11)

Configuration:

• eth0 10.1.50.2/24, gateway 10.1.50.1 (VLAN 50)

• eth1 10.1.60.2/24, gateway 10.1.60.1 (VLAN 60)

• eth2 10.1.70.2/24, gateway 10.1.70.1 (VLAN 70)

Problem:

UniFi OS installs one default route per interface, with no metric. Without local changes, "ip rule" shows only the stock rules, and:

ip route get 10.1.20.3 from 10.1.60.2

-> 10.1.20.3 from 10.1.60.2 via 10.1.50.1 dev eth0

A client at 10.1.20.3 connects to 10.1.60.2:445. The SYN arrives on eth1, and the SYN-ACK leaves eth0 with source 10.1.60.2. The UDM-Pro receives it on VLAN 50 and does not forward it, so SMB from a different VLAN either times out or fails to mount. Same-subnet clients are unaffected.

The UI provides no route priority, primary interface or per-interface reply setting that I can find. A blank gateway fails validation in the UI, and setting gateway to 0.0.0.0 seems to have no effect on the issue. Local source-based routing restored service (cold SMB connect fell from 9-15 s to 0.04 s), but it is not a supported configuration.

Questions:

1. What is the supported configuration for using these ports on separate routed VLANs?

2. Can you add per-interface route metrics or a primary interface, allow ports with no gateway and no default route, or reply through the receiving interface?

3. Until then, is a systemd-networkd drop-in supported, and will it survive updates?

4. Does the UNVR G2 use the same network behaviour?

@UI-Team, full support files for the ENAS, UDM-Pro, and UNVR G2 are attached.

Just a slight issue: let me choose the exact destination path for UNAS backups.

Setup: UNAS Pro, UniFi Drive 4.4.20. Backup task "Local Backup" → Source "Rsync Server" (Transfer Encryption: Over SSH), two modules selected, destination Shared Drives > misc, mode "Skip Existing". Works well and fast.

Problem: Drive always writes to <destination>/<task name>/<module>/…. The task name is a label, yet it decides where data is stored. Renaming the task changes the path, so the next run downloads everything again. While choosing a name, the same 11 GB were downloaded three times. Files can't be placed into an existing library (e.g. media) without an extra folder level.

Requests:

1. An option to sync a module's contents directly into the chosen destination folder (like rsync src/ dest/), or a per-source destination mapping (module tv → Medien/TV, module film → Medien/Filme).

2. Decouple the task name from the storage path. Renaming a task should never move or re-download data.

3. Schedules more frequent than daily (hourly or every N minutes). Right now only daily, weekly and monthly are offered.

4. Optional "only files that are new since the last run" mode. With "Skip Existing", files that were moved or deleted at the destination are downloaded again on the next run.

5. Document the "Over SSH" mode (module list from ~/rsyncd.conf on the source) in the help center. The article "Troubleshooting UNAS Backup Issues" currently says only daemon mode is supported.

Thanks!

I have UNAS 8 Pro with management via copper interface with DHCP IP.

I need to setup static IP without gateway (storage only LAN) on second 10G interface.

System doesn't allow me to do that. The same apply for DNS.

How can I setup network without gateway and DNS - NAS have them on Port 1, works ok.

Why ? This is typical L2 network just for storage access and not having gateway there is security requirement.

It has been quite some time now, I'm not sure what is the holdup?

We want to sell this ASAP to customers!

I'm considering purchasing a UNAS Pro (original, 7 bay one) and a rack to put it in (amongst other things). The UNAS Pro specifies for rack depth: "Posts depths ranging from 650 to 1,000 mm (25.6–39.4”)"

That raises two questions for me:

1. Is there actually a maximum rack depth, given the UNAS Pro mounts on ears and not rails?

2. Can I put it in the Ubiquiti 12U rack (standard depth) with a depth of 600mm"?

By the specs the answer to 2 is no, but the depth of the machine is only 325mm". That leaves a whopping 275mm (11") behind the unit in the 12U rack, and 325mm (13") by the stated minimum rack depth!

So am I going crazy that the minimum post depth as specified for the UNAS pro seems wrong? If wrong, what is the actual minimum rack depth I should be considering when shopping for a rack?

Thanks!

Support ticket: #6087504

Setup

- UNAS Pro 8, UniFi OS 6.0.11 (updated from 5.1.33 on 18 Oct)

- Storage Pool 1: RAID 6, 5x Seagate Exos X16 16TB

- Single PSU

What happened

- 20 Oct (two days after the 6.0.11 update): the drive in one bay (ata3/sdc) started repeated SATA link failures: "hard resetting link", "link is slow to respond", link renegotiating down to 1.5 Gbps / UDMA/33, and "retrying FLUSH 0xea". Hundreds of resets every night, mostly between 02:00 and 05:00.

- The same day, Btrfs began logging "parent transid verify failed", i.e. metadata writes that never reached disk.

- The pool kept running for 6 days with no alert I noticed.

- 26 Oct, after a normal reboot, the pool fails to mount: transid failure on the checksum tree → "failed to read root (objectid=7)" → open_ctree failed. All four superblock backup roots point to the same bad block.

- UI now shows "File System Issue", "2 / 5 HDDs" (md shows [5/5] [UUUUU]), and Drive offers to set up storage from scratch.

What's healthy

- The drive in that bay has clean SMART: 0 reallocated, 0 pending, 0 uncorrectable, 0 UDMA CRC errors. So this looks like the bay, backplane, power or link handling, not the disk.

- md RAID 6 is healthy, LVM is intact, and btrfs restore can read almost all the data. I'm copying it off over a direct link now, using read-only tools only. No repair, no resync, no setup/reformat.

Questions for the team

1. Is there a known issue with SATA link handling or power management in 6.0.11 on the UNAS Pro 8? The resets began two days after the update.

2. What exactly does the new 6.0.11 "on-demand file system repair" run, and is it safe for a transid/csum-tree failure where one drive may hold stale blocks? I haven't run it.

3. Why did the pool keep running for 6 days without alerting me, given hundreds of link resets and Btrfs transid errors?

4. Can engineering do an in-place recovery (e.g. reading the array without the suspect drive so its stale blocks rebuild from parity) instead of recreating the pool?

Happy to provide more logs. Anyone else on 6.0.11 seeing "hard resetting link" or transid errors in dmesg?