Release LoadGen FrontEnd: 1.0.0.12422
LoadGen FrontEnd
[#9552] Offline update bundle — update an air-gapped appliance from one signed artefact — the Updates tab gains a path for appliances with no route to the internet. Stage a signed bundle from the appliance's intake directory, let the appliance verify it and read its own update plan, then apply it as one recoverable transaction covering the agent, the host packages, every container image, the binary archive and the release catalog. The panel follows the run through the agent replacing itself and through a reboot, and a trust line states which release identity the appliance accepts. Requires the API from this release (1.0.0.12421) or later; against an older API the panel renders a single grey line saying so, with no error banner and no dead controls.
Repository health — scan your repositories for outdated packages, vulnerabilities and end-of-life runtimes — Uptime → Checks can now scan the repositories of an Azure DevOps organization or project, a GitHub organization or user, or a GitLab group, and report outdated packages, known vulnerabilities and end-of-life runtimes. The wizard builds its settings from the check catalog itself, so the source, scope and rules for each provider are offered as proper selects, checkboxes and toggles rather than free text, and a new Repositories report view presents the findings. Requires the API from this release (1.0.0.12421) or later; on an older API the repository check types simply aren't offered.
Import uptime checks from a JSON document — Uptime → Checks gains an Import button (and an Import from JSON link on the empty state) leading to a wizard: drop a
.jsonfile or paste it into the editor, review what was found, move every inline password and key into Credential Manager, then import in one go. You can also generate an Empty or Example template for any check type to start from. Requires the API from this release (1.0.0.12421) or later; on an older API the page opens and explains what it needs instead of failing.Credential profiles now record expiry and the account they belong to — a credential profile can say when its secret expires and which account it was issued for, and the Credential Profiles list shows both in new Expires and Identity columns. An expiring credential reaches you through the Reminders hub and a sign-in nudge, so a token with a long life no longer runs out unannounced. Requires the API from this release (1.0.0.12421) or later; against an older API the Expires column reads Not tracked and everything else behaves as before.
[#9614] Fixed the Updates tab's Frontend tile reporting Updates available with no working way to act on it, while the Appliance Agent, API Container and MCP Server tiles all offered a clearly styled button. The tile's button was rendering white on white rather than being disabled; all component tiles now paint their accent colour, so the Frontend update button is visible and clickable like the rest.
[#9618] Fixed the Uptime analytics panel presenting a capped figure as a measurement of the selected period — expanding a check and choosing 7d reported Total Checks 1,000 with the passed and failed split adding up to exactly 1,000, whatever the period actually held. Totals now cover the whole selected window. In the case where the API cannot reach a datasource and has to answer from a shorter buffer, the panel shows an information strip above Response Time Details saying the figures may be short — the numbers stay visible and legible, but they are no longer presented as a full-period measurement.
[#9616] Fixed the Package drift card on the Launcher Appliance detail page staying stale after an operator action — an appliance that had just been rebooted could still read Reboot required — pending since 15d ago a minute and a half later, even though the platform already held a correct inventory. Recompute inventory and the post-reboot watch now both keep checking until the appliance's inventory actually moves and update the card on screen, instead of promising that the numbers will refresh shortly and never fetching them again.
[#9615] Fixed Test Runs keeping a selection the table no longer showed: after changing a filter, the Delete (N) count went on offering to delete runs that were no longer visible — the reported case showed Delete (12) above Showing 1-1 of 1 tests. Changing or clearing a filter now clears the selection. Paging still keeps it deliberately, since the pages are one result set seen through a different window and selecting runs across pages is how Compare and bulk delete are meant to be used.
Fixed pages going blank behind an "An unhandled error has occurred." banner. Monitoring → Schedules dropped the connection whenever the displayed week contained a scheduled run, and Global Variables did the same. The same markup defect is corrected on the Insight Additional Monitoring, Registry and System Monitoring tabs and on SessionSight → Config, and a build-time guard now prevents it recurring.
Fixed the Uptime → Scripts list printing the literal text v@script.Version in its Version column on every row instead of each script's version number.
Fixed Uptime → Checks → New showing "Could not load catalog" with no check type offered at all, leaving Custom (advanced) as the only way to define a check. A catalog entry that takes the general interval or timeout no longer costs you the entire list of check types.
Fixed clicking a check tile on a saved Uptime → Dashboards board opening the Not Found page instead of the check.
Fixed the load profile wizards ignoring the technology and step in a link when you were already on a wizard page, so opening Local Client or Test Client from a link showed the technology from the previous visit rather than the one asked for.
The Credential Profiles page and the credential pickers in the wizards no longer come up empty when the API offers a credential type this build doesn't recognise. An unfamiliar type now degrades to a single clearly marked entry instead of discarding the whole list.