Unofficial

StorageDiag · Workbench & Utilities

Free forever

Engineered to analyze. Built to maintain.

Unlock the raw mechanical reality of your QNAP NAS. StorageDiag Workbench & Utilities gives you direct access to CPU steal metrics, SMART attribute history, and block-level disk fragmentation—giving you absolute mastery over your environment.

— The ultimate Diagnostic Suite. Legacy: Silas Diagnostic Utilities

Watch the full reel

Recorded on a live NAS — every window, every number, real.

Stop guessing. Start knowing.

Pulse: twelve per-core meters over a filterable tree of 2004 live processes, each with CPU, memory, RSS and thread count Pulse CPU tab: per-core busy, iowait, steal and current clock against the governor in force Pulse History: six hours of CPU busy recorded on the NAS, two load spikes standing out of the baseline
Live · every process, every corePulse Advanced Process & Resource Monitor The ultimate live-action dashboard for your QNAP. Pulse replaces standard CLI tools like top, htop, iotop, and netstat with a unified, high-density interface. It maps exact hardware resource consumption to specific processes, including vCPU tracking for virtual machines and containers. Key capabilitiesLive tracking of CPU steal, memory swapping, and exact thread usage.Direct mapping of network sockets to the processes behind them.Real-time application database monitoring and live log tailing in one view.
See it running Download
BlackBox: CPU busy, blocked tasks and CPU frequency stacked on one 24-hour axis BlackBox heat map: a week of temperatures, one lane per sensor, hot spells standing out as bands
Flight recorder · the past, on a timelineBlackBox The NAS Flight Recorder High-resolution historical telemetry. Rewind and find the crash. BlackBox records system metrics continuously—capturing CPU IOWAIT, thermal profiles, disk latency, and RAM usage—without flattening crucial data spikes into meaningless averages. Whether an issue happened two minutes or two months ago, BlackBox lets you scrub through the timeline to see exactly what triggered a system bottleneck or crash. Key capabilitiesMinute-by-minute historical data retention (hours, days, weeks, months).Deep analysis of core-specific temperatures and CPU IOWAIT spikes.Visual heatmaps for instant identification of recurring anomalies.
See it running Download
DriveTime Health: the fleet rail beside one drive history, a Seagate 10 TB with 369 reallocated sectors marked critical DriveTime Speed: a read-only scan of a storage pool volume ending in a named weak link — slot 14, 369 reallocated sectors, G-sense during the run DriveTime Disk Map: the free-space map of a volume beside the most fragmented files, with the run count for each
Predictive · every bay, over timeDriveTime Analyzer Predictive Disk & Array Diagnostics Beyond basic SMART. See the actual mechanical health and layout of your drives. A “PASSED” SMART status means nothing if a drive is silently dying. DriveTime tracks drive attributes (like reallocated sectors and temperature) over time as a graph, letting you spot hardware degradation before a RAID failure occurs. It also profiles active I/O to pinpoint exactly which single disk is dragging down your entire storage pool. Key capabilitiesTime-series graphs for critical SMART attributes (track the climb as well as the current state).Live I/O latency monitoring (Disk Await) and sequential read bandwidth testing.Block-level free space and fragmentation mapping for both EXT4 and ZFS volumes.
See it running Download
NetScope Packets: a bounded live capture beside a decoded frame — Ethernet, IPv4 and TCP expanded next to the hex NetScope: every device found on the local segments, with name, vendor, address and whether it answered NetScope DNS: the gateway resolver measured against Cloudflare, Quad9 and OpenDNS, answer by answer NetScope Bandwidth: an iperf3 client with the NAS peers it discovered, duration, direction and protocol NetScope VPN link: an end-to-end check of a tunnel — MTU, round trip, jitter and loss, with a plain verdict
Deep inspection · both ends of the linkNetScope Deep Network Packet Inspection Stop blaming the network. Prove it with exact traffic data. NetScope brings Wireshark-style packet capture and dissection straight to your browser. If SMB shares are dropping, links are negotiating poorly, or DNS is failing, NetScope captures the raw traffic and pinpoints exactly where the connection breaks—whether it’s the client, the switch, or the NAS. Key capabilitiesLive packet tracking, link path analysis, and resolver diagnostics.One-click throughput testing (iperf integration).Identifies dropped packets, retransmissions, and protocol-specific bottlenecks.
See it running Download
CPU Lab: a finished run — the CPH index, every kernel measured single-threaded and across all cores, with scaling and wait CPU Lab mid-run: the kernel table filling in while the core map shows which logical CPUs are loaded CPU Lab before a run: the kernels it will measure and what each one stresses — cache, DRAM latency or SIMD
Benchmark · sustained, not a burstcpuhound Sustained Performance Benchmarking Honest computing metrics. Know exactly when and why your CPU throttles. cpuhound goes beyond synthetic scores by applying sustained computing loads to your QNAP’s processor. It monitors thermal envelopes, cache latency, and thread scaling until the performance settles, revealing the true operational limit of your hardware before thermal throttling kicks in.
Open the board Download
memhound: the arena map mid-test with failing cells marked red, and the bit-flip log naming the module and address memhound verdict: the deep test finished and failed, with the faulty module identified by slot memhound Benchmark & Quality: the qualification verdict for the memory in this machine, measured against its JEDEC rating
Burn-in · the DIMM that liesmemhound Live RAM Doctor & Burn-in Tester Enterprise-grade memory diagnostics without taking the NAS offline. memhound brings memtest86-class rigorous pattern testing directly to your active QNAP environment. It safely reclaims RAM (including the ZFS ARC) to perform deep stress tests, sweeping memory cell by cell to detect transient bit flips, measure true bandwidth, and verify ECC functionality. Key capabilitiesLive-system memory testing with zero false positives.Measures memory bandwidth (GB/s), idle/pending latency, and structural integrity.Option to create a bootable USB for isolated, bare-metal hardware validation.
See it running Download
Dr. Trace: kernel messages and system events merged onto one timeline, severity-tagged, with the incident rail beside them Dr. Trace with Focus on: the same log cut down to the critical events worth reading
Forensics · one merged timelineDr. Trace Merged Log Timeline Every log the NAS keeps, on one timeline. Dr. Trace merges the kernel ring buffer with the system event log and puts them on a single ordered timeline, tagged by severity. Focus strips the routine chatter so what is left is the sequence that actually explains the incident. Key capabilitiesKernel messages and system events ordered together, not read side by side.Focus cuts a hundred thousand lines down to the ones that matter.Works on a diagnostic bundle somebody sent you, or on the live machine.
See it running Download
Samba: every share with its volume, permissions and recycle-bin state, over the service status band Samba Access: the matrix of which principal reaches which share, and with what rights
Audit · who can reach whatSamba Shares & Access Audit See every share, and exactly who can reach it. The Samba panel reads the share definitions, the ACL tokens and the service settings together, then answers the question a support case actually asks: who has access to what, and is anything reachable that should not be. Key capabilitiesA principal-by-share matrix built from the real ACL tokens.Advanced folder permissions, Windows ACLs and ACL 2.0 reported from the keys that set them.Flags a share reachable by guests, and an SMB service that is enabled but not running.
See it running Download
Virtual machines and containers: every guest with its vCPU allocation, memory, firmware, disks and network, against the host total
Hypervisor · vCPU against pCPUVirtual Machines Guests & Containers Look inside every guest the NAS is hosting. One window for the virtual machines and containers on the box: the vCPUs each one holds against the host total, memory, firmware and boot mode, the disks behind them and the networks they sit on. Key capabilitiesvCPU allocation against the host, so over-subscription is visible, not inferred.Disks, buses and snapshot mode per guest.Containers and virtual machines in the same inventory.
See it running Download
Scheduled · DriveTime and memhound, unattendedHW Vital Tests Daily Disk & RAM Tests The NAS tests its own disks and memory every day. By default, every disk except USB drives gets a quick DriveTime surface test at 04:00, one disk at a time. At 05:00 memhound tests the memory that is free at that moment, and every service keeps running. Once every three weeks, on a Saturday, a deep memory test runs in its place: it stops all apps and file sharing until it ends, then starts them again. Faults go to the QTS System Logs, and the window holds the schedule and the recent results. Key capabilitiesDisks: a 1 MiB check in each of 4096 equal parts of the disk; on SATA disks the verdict also tells a failing disk from a bad slot, cable or backplane.RAM: memhound leaves 256 MiB free, re-checks free memory before every block, and is the first process the kernel kills if memory runs out.On by default; with Virtualization Station installed, the deep test stays off until an administrator allows it.
Download

Install it on a healthy machine

A flight recorder is worthless the day you install it and priceless a month later — the fault you will want to look up has not happened yet. One QPKG, App Center → Install Manually. Not sure which build? Run uname -m on the NAS.

x86_64Intel / AMD · most TS, TVS, TES ARMv8 · aarch64TS-x28A, TS-x33, TS-x35 ARMv7 · armhfTS-x31P, TS-x31X, TS-x31K

NetScope for Windows — the other end of the link QTS 5.2+ & QuTS hero 6 · admin-gated · nothing leaves the NAS

Pricing

everyone has one now

So, like everywhere else: here is the price list.

Pick a plan. Both are the same price, and neither one asks for an e-mail, a licence key or a credit card. One of them gets you every instrument on this page.

Plan A

I’m not interested

No cost

  • Read the page, close the tab
  • Nothing to install, nothing to unsubscribe from
  • No account, no tracking, no cookie banner
  • Come back the day the pool will not mount

Keep reading

Plan B · recommended

Download as QPKG

$0 USD · forever

  • Freeware — no price, no licence key, no account
  • Install on your own NAS — QTS 5.2+, QuTS hero, three architectures
  • Install on as many servers as you own — no seat count

Download for free

“Why won’t my RAID pool mount? And why is dmesg full of random call traces?”

I have been asked that more times than I can count. The answer, more often than anyone likes, is one failed RAM module. Data moving between the CPU, the PCIe lanes and the disks was being corrupted in both directions — which means the corruption was already on the platters before anyone noticed the pool. A memory tester and a CPU/memory benchmark are not luxuries. They are the first thing to run, and the thing nobody runs.

Did you know your CPU may be running slower than the number on the box? The moment a core reaches 90 °C it throttles its clock down and stays there. Better cooling gives you back efficiency you already paid for — but only if something shows you the clock and the temperature on one line, over time.

Add vCPU oversubscription on virtual machines, I/O delay warnings, blocked tasks, SMART attribute history, block-level fragmentation and the rest, and you have a picture of your hardware that the vendor’s dashboard was never built to show. These are basic instruments. Basic instruments should not cost money.

1 DIMMis all it takes to corrupt a pool silently. memhound finds it before the array does.
90 °Cand the clock drops. BlackBox draws the throttle next to the temperature that caused it.
8 : 32physical cores to vCPUs is the usual home hypervisor. Pulse shows which guest is stealing from which.

Questions people actually ask

FAQ
Can BlackBox show me what happened last night? Or last month?

Yes. BlackBox is the flight recorder: it keeps CPU busy, every core’s clock, the temperatures, blocked tasks, memory pressure and a few other things as graphs you can scrub through — hours, days, weeks back. Install it on a healthy machine; the fault you will want to look up has not happened yet.

Is this an alternative to Zabbix, Observium, Prometheus, Grafana…?

No. Those watch a fleet from the outside and page you when a number crosses a line. StorageDiag looks at one machine, from the inside, close up — the drive head, the memory module, the core that is throttling — and explains what it sees. Run both if you like; they do not compete for the same job.

What is vCPU oversubscription, and why does my NAS crawl when the VMs are busy?

It is the most common performance issue on a home hypervisor, and almost nobody measures it. A NAS with 8 physical cores and four virtual machines with 8 vCPUs each: the guests believe they own 32 cores, the host has 8, and every guest waits on the others. That is why BlackBox exists — it puts oversubscription and CPU temperature next to the delayed I/O and wait charts, so you can see how much of the slowdown is yours to fix.

What does it cost, really? What is the catch?

Nothing, and there is none. One QPKG, free forever, no licence, no account, no data leaving the NAS. The about page shows how to support it.

Download for free now!