UGREEN NAS Troubleshooting: Find the Performance Bottleneck

When a UGREEN NAS feels slow, test one large file over a wired local connection before changing RAID settings, installing SSD cache, or buying new hardware. That result establishes whether the bottleneck is the network, storage pool, or workload.

Around 100 to 110MB/s is normal for a healthy 1GbE path. If one large file transfers quickly while a folder containing thousands of small files remains slow, the network is probably working correctly.

Protect the data before testing performance. If Storage Manager reports a degraded pool, S.M.A.R.T. errors, repeated read failures, or a disconnected drive, stop benchmarking and secure an independent copy of important files. Clicking or grinding sounds also require immediate attention.

Identify the Symptom Before Changing UGREEN NAS Settings

Start with what the NAS is doing. Different symptoms require different tests.

Run one large wired local transfer
                │
        Below link-speed range?
          ┌─────┴─────┐
         Yes           No
          │             │
Check network link,   Check pool status,
SMB security, MTU,    background tasks,
cable, and switch     memory, and workload
          │             │
      Still slow?    Slow only under load?
          │             │
Check storage path   Isolate the application
and client device    or concurrent users

Test UGREEN NAS Transfer Speed Under Controlled Conditions

1. Connect the computer and UGREEN NAS through wired Ethernet on the same local network.
2. Access the UGREEN NASync through its local IP address.
3. Pause optional cloud synchronization, backups, downloads, media scanning, and photo indexing.
4. Use SMB to copy one file several gigabytes in size.
5. Test both uploading to the NAS and downloading from it.
6. Repeat the transfer using a different Ethernet cable and switch or router port if the result is unexpectedly low.

Avoid using a browser upload, Wi-Fi, remote access, or a folder containing thousands of files for the initial test. Each introduces another variable.

For a complete step-by-step procedure, follow our slow NAS transfer troubleshooting guide.

Network Bottlenecks: Link Speed, SMB, and MTU

Every device between the computer and NAS must support the intended speed. One 1Gbe port, damaged cable, limited USB network adapter, or incorrectly configured switch can slow the entire path.

A single HDD or a small HDD array may reach its own limit before the network reaches 10GbE capacity.

In UGOS Pro, open Control Panel > Network > Network Connection and check the negotiated speed of each active port. If a 1GbE, 2.5GbE, or 10GbE connection has fallen back to 100Mbps:

• Replace the cable with a known-good CAT6 or better cable.
• Move the connection to another router or switch port.
• Bypass USB network adapters, docks, powerline adapters, and intermediate devices.
• Confirm that the computer’s adapter negotiated the intended speed.
• Restart the affected adapter or switch port after replacing the cable.

Also review network data-control rules. A bandwidth limit assigned to a service, interface, or port can cap transfers even when the physical connection is healthy.

SMB Signing and Encryption Can Shift the Bottleneck to the CPU

SMB signing protects message integrity. SMB encryption protects both integrity and confidentiality. Those protections require cryptographic processing on the client and NAS.

Microsoft identifies both features as possible SMB performance factors and notes that the effect depends on CPU capabilities and available processing time. Starting with Windows 11 24H2, SMB signing is required by default. Do not disable signing or encryption merely to improve a benchmark. They protect against tampering, spoofing, relay attacks, and eavesdropping. Microsoft’s SMB performance guidance explains the tradeoff.

If the network and storage pool test normally but SMB throughput remains low, watch CPU usage on the NAS and computer during the transfer.

On Windows, PowerShell can display the SMB dialect and security state of active connections:

Get-SmbConnection | Select-Object ServerName, ShareName, Dialect, Signed, Encrypted

If CPU utilization reaches its limit during a signed or encrypted transfer, the processor has become the bottleneck.

Reset Jumbo Frames Before Diagnosing Intermittent Transfers

Jumbo frames work only when the NAS, client, switch, and every intermediate device use a compatible MTU. A mismatch can allow a transfer to begin and then stall when larger packets cannot pass correctly.

Work through it in this order:

1. Return the NAS, client, and switch to the standard 1500-byte MTU.
2. Repeat the controlled large-file test.
3. Enable jumbo frames only if every device along the path supports and uses the same MTU.
4. Retest after each change.

Jumbo frames are not required for 2.5GbE or 10GbE. They can reduce packet-processing overhead, and a consistent 1500-byte configuration is better than an inconsistent 9000-byte one.

When the NAS Keeps Disconnecting

Intermittent access commonly points to a cable, switch port, client adapter, IP-address change, MTU mismatch, or link-aggregation problem.

Test the NAS through one physical network port with link aggregation disabled. If the connection becomes stable, inspect the bond mode and switch configuration before enabling it again.

If the local IP address has changed, saved network-drive paths and bookmarks may still point to the old address. Confirm the current address in UGOS Pro and update the connection on the client.

When SMB works normally on one computer but NFS or another client behaves differently, compare the client and protocol configuration before replacing hardware. Our UGREEN NAS file-sharing guide explains SMB, NFS, and shared-folder permissions.

Link Aggregation Increases Total Bandwidth Across Connections

Link aggregation combines compatible NAS network ports for greater aggregate bandwidth, connection redundancy, or both. It is most useful when several clients access the NAS at the same time.

Link aggregation alone does not turn one 1GbE connection into a 2GbE transfer. A single TCP flow normally uses one physical link. Multiple clients or multiple supported flows can share the combined bandwidth.

UGREEN DH series models do not support link aggregation. On supported models, UGOS Pro provides several bond modes:

To configure it, open Control Panel > Network > Network Connection > Link Aggregation.

All participating ports must connect to the same switch and VLAN. Bond0, Bond2, and Bond4 also require matching configuration on a compatible switch. Our guide to choosing a NAS network switch covers port speeds and managed-switch features.

Creating a bond briefly interrupts network access. Firewall and traffic-control rules attached to the original interfaces must be recreated for the bonded interface.

Some consumer routers can interpret Bond1 or Bond6 traffic as an IP conflict. If the NAS becomes unstable after aggregation is enabled, remove router-side IP-to-MAC binding, test one physical port, and confirm that the selected mode matches the switch.

Storage Bottlenecks: Pool Activity, Available Space, and Drive Health

If the controlled network test is normal, check what the storage pool is doing.

Open Storage Manager and review:

• Storage-pool status
• RAID synchronization or rebuilding
• Volume optimization
• Remaining free capacity
• Drive temperatures
• S.M.A.R.T. results
• Disconnected or repeatedly reconnecting drives

RAID synchronization, rebuilding, and volume optimization generate sustained disk activity. File transfers, applications, and media indexing remain slower until that work finishes. Keep the NAS powered and avoid interrupting the operation unless UGOS Pro reports an error.

Next, pause optional workloads one at a time:

• Cloud synchronization
• Computer backup jobs
• Photo indexing and AI analysis
• Media-library scanning
• Downloads
• Containers
• Virtual machines
• Antivirus scans

If performance returns after one task is paused, the problem is resource contention.

A Nearly Full Pool Needs Working Space

If the storage pool is approaching full capacity, free space before evaluating performance. File-system allocation, metadata, snapshots, and parity writes all need working room.

Drive Hibernation Causes a Normal First-Access Delay

If only the first folder or file takes several seconds to open after a long idle period, the HDDs may be waking from hibernation. Subsequent access should return to normal speed.

Disabling drive hibernation removes that wake-up delay, but it keeps the drives spinning and increases power consumption and noise. Leave it enabled when the short first-access delay is acceptable.

Read the SMART Details, Not Only Pass or Fail

When detailed S.M.A.R.T. attributes are available, pay particular attention to these values:

An increasing CRC error count points toward the drive-to-NAS connection rather than the Ethernet cable. Reallocated and pending sectors point more directly toward the drive media. Seagate’s SMART guidance documents these attributes.

A single historical value does not always prove active failure. A rising count, repeated disconnections, read errors, or several warning attributes together require action.

If a drive reports errors, follow the safe sequence in our guide to replacing a failing NAS drive.

UGOS Pro Memory Compression: What High Memory Usage Means

Memory compression reduces the amount of data UGOS Pro needs to move from RAM to slower disk-based swap space. When memory pressure reaches the configured threshold, UGOS Pro compresses part of the data held in memory.

The setting is available at:

Control Panel > Hardware & Power > General > Memory compression

Compression can improve responsiveness when the NAS is short on available RAM, but compression and decompression require CPU time. The result depends on the workload.

High displayed memory use alone does not prove that the NAS needs more RAM. Check whether:

• Applications are becoming unresponsive.
• Swap activity remains high.
• CPU usage rises sharply while compression is active.
• Containers or virtual machines exceed their assigned limits.
• Performance improves when a memory-heavy application is stopped.

SSD cache cannot substitute for system memory. Cache accelerates suitable storage activity; it does not provide working memory to UGOS Pro or its applications.

SSD Cache Performance Depends on the Workload

SSD cache provides the clearest benefit when the NAS repeatedly accesses the same small blocks of data. Suitable workloads include:

• Frequently opened shared files
• Databases
• Virtual-machine data
• Application data
• Multi-user folders with repeated access patterns

Large sequential transfers behave differently. Copying a video file once, streaming media, or reading a large archive touches the data in order and may not reuse it soon. SSD cache therefore has little opportunity to help.

For sustained large-file performance, the more relevant limits are:

• Network speed
• HDD or SSD pool speed
• RAID configuration
• Number of active drives
• Client storage performance

If testing confirms a hardware ceiling, use our guide to improving NAS read and write speed to determine what needs to change. If the existing port speed, drive count, memory capacity, or storage configuration cannot support the workload, you can browse and filter NAS Storage models by network speed and bay count.

Multiple Users, Containers, and Remote Access

A NAS can perform normally for one user and slow down when several workloads run together. Each user, container, backup task, media transcode, and virtual machine competes for network bandwidth, CPU time, memory, and disk IOPS.

Watch for these patterns:

• One user is fast, but several users are slow: aggregate network bandwidth or disk IOPS may be exhausted.
• Large files remain fast, but applications respond slowly: latency, memory, or CPU demand is more likely.
• Performance drops only during media playback: check whether the NAS is transcoding.
• The NAS slows after a container starts: review that container’s CPU, memory, and storage activity.
• Local transfers are fast, but remote access is slow: test the internet upload speed at the NAS location and the download speed at the remote client.

The detailed fixes for shared workloads belong in our guide to improving NAS performance for multiple users. If the slowdown occurs only outside the local network, review the factors affecting NAS performance for remote work.

Before Contacting UGREEN Support

Record the following information before contacting UGREEN Support:

• UGREEN NAS model
• UGOS Pro version
• Client operating system
• Client network-adapter model and negotiated speed
• NAS port and negotiated speed
• Router or switch model
• Ethernet cable category
• MTU configured on the NAS, switch, and client
• Connection method, such as SMB, NFS, app, browser, or remote access
• SMB dialect, signing state, and encryption state
• Test-file size
• Upload and download results
• Storage-pool status and remaining capacity
• Active synchronization, backup, indexing, container, or virtual-machine tasks
• Drive S.M.A.R.T. warnings
• Relevant screenshots, timestamps, and error messages

A report such as “a 10GB local SMB transfer reaches 104MB/s over 1GbE, but remote downloads remain below 5MB/s” identifies the affected part of the path. “The NAS is slow” does not. A controlled result lets our support team investigate the actual bottleneck without repeating the entire diagnostic process.