The System Dashboard is where you go to see how your server is actually doing. It is admin-only, it lives right at the top of the sidebar under Dashboard > System Dashboard, and it is the one page that answers "is the box healthy, and is there anything I need to do?" This release is about making it answer that question better. There is a new panel that reads your hardware straight from the kernel, the whole page is faster and easier to read, it finally behaves on a phone, and the two places you go to keep the server current - the updates table and the changelog - both got a lot smarter.
Hardware Health, at a glance
The headline is a brand-new Hardware Health panel. It sits on the System Dashboard, just below the RAID panel, and it gives you a one-glance read of the physical box - the parts that never showed up in a dashboard before.
It is laid out in three parts. Platform shows your BIOS, board, and product identity. CPU shows temperatures, fans, and throttling, how many cores are actually online, your microcode revision, the CPU's published vulnerability status, and a count of any kernel faults grouped by which core they landed on. Memory shows ECC error counters, whether the kernel has flagged any corrupted memory, and a full inventory of your installed modules against what the system can see.
Two things make this panel different from the usual "install a monitoring agent" story. First, it needs nothing installed. Every label and threshold it shows is read straight from the kernel's own interfaces, so there is no package to provision and no daemon to babysit. Second, it is honest about what it cannot know. On a virtual machine, where firmware and temperature readings are meaningless, the panel says so plainly instead of inventing a fake green light. A reading it cannot take is shown as "not measurable," never as a pass.
The panel gives you a single status - healthy, warning, or critical - rolled up from everything it measures, with the specific findings called out so you know exactly what tripped it. Findings you have seen and understood can be hidden or dismissed so the panel stays meaningful, and where a finding has a real reset - clearing the history of kernel dumps, or acknowledging a past thermal-throttle event - that action is right there. There is also an admin control to set the CPU power profile without leaving the page.
This grew out of a real diagnosis where a single failing CPU core could only be found by hand-mining kernel logs over SSH. Now it is a panel. The full reference, including every field and how the status is derived, is in the System Monitoring docs.
A System Dashboard that works everywhere
The rest of the page got the polish it deserved. We went through the System Dashboard surface by surface and fixed the things that quietly made it harder to use.
It reads cleanly on a phone now. The page no longer scrolls sideways on a small screen, labels stop getting cut off mid-word, and the panel headers stack sensibly instead of crushing a title under a row of buttons. Controls are sized for a thumb. If you have ever tried to check on your server from your phone and given up, try it again.
It is easier to read at every size. The tiles and cards now size themselves to the space they actually have rather than to a guess about your screen, so a label reads as "This month" in full instead of "MONT..." with the sidebar open. The status colors across the health panels were tuned to meet the readable contrast floor, so a warning or a critical reads clearly on every theme, light or dark. And the Diagnostics panel now opens ready to use, with the application details already in view, so copying a snapshot for support is one glance instead of two clicks.
None of this is flashy on its own. It is the kind of work you only notice when it stops getting in your way. If you want the full tour of everything the page can do - live metrics, health scoring, bandwidth history, RAID, and the telemetry history store - it is all written up in the System Monitoring docs.
Every update in one place
The Updates Available table on the System Dashboard used to show only your APT package updates. Now it shows everything waiting for you, together. Alongside the APT rows, it lists QuickBox's own two components whenever an update is actually available: QuickBox CLI, the core command-line engine, and QuickBox Dashboard, the web interface you are looking at. Each row shows what you have installed against what is available.
A Source filter lets you narrow the table to just QuickBox or just APT when you want to focus. The two QuickBox rows carry an Open Updates link that takes you straight to the changelog, and they show their command for reference too - qb update quickbox for the core and qb update dashboard for the interface - so it is clear what each one does. The Update APT Dependencies button applies to the APT rows only, and it says so, so there is no confusion about what it will and will not touch.
The whole thing degrades gracefully: if the version check cannot reach out for a moment, the table simply shows your APT rows rather than erroring. The full write-up of the updates workflow lives in the Updates and Changelog docs.
A changelog that knows Core from Dashboard
The changelog page - System > Changelog in the sidebar - is where you read what an update actually changes. It got two real improvements this release.
First, it finally tells Core from Dashboard. The QuickBox CLI (on the 3.x line) and the Dashboard (on the 4.x line) are separate components on independent version lines, and the page now treats them that way. Each has its own "current" and "latest" markers measured against its own installed version, and a pending Core release is never mixed in with a pending Dashboard release. The combined view sorts by real release date, so Core and Dashboard entries interleave in the order they actually shipped instead of one component stacking on top of the other and the other scrolling off the bottom.
Second, and this is the one you will feel if your install is a few versions behind: the page now shows you what is coming before you take it. When you are behind, it gathers the release notes for every version between the one you have installed and the latest, and groups them at the top under Available in your next update with a badge for how many are waiting. You can read exactly what each pending release changes before you download and apply anything - no more updating blind. When you are fully current, the group simply is not there and the list stays flat.
The complete guide to the page, including previewing pending updates and the CLI equivalents, is in the Updates and Changelog docs.
Where this goes next
The System Dashboard is the page you open when something feels off, and it should give you a straight answer fast - on your desk or on your phone. That is the bar this release sets, and it is the bar every panel we add from here has to clear. If something reads too faint, breaks on your screen, or tells you less than it should, tell us. That feedback is how this release happened, and it is how the next one gets better.