Resolves: #34771
This moves away from relying on discontinued bitnami charts and instead
adds a small mysql chart, a valkey/redis chart and a brief guide update
on how to migrate from one to the other.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Helm chart bumped to v7.0.0.
* Replaced Redis with Valkey as the caching backend and added Valkey
configuration options.
* Added an optional embedded MySQL chart with configurable auth,
persistence, service, and credentials handling.
* **Chores**
* CI now adds the Valkey Helm repository and builds chart dependencies
before templating.
* .gitignore adjusted to only ignore packaged chart archives (*.tgz).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: georgekarrv <1501415+georgekarrv@users.noreply.github.com>
- Enable `standard` RDS database performance insights
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Release Notes
* **Chores**
* Enhanced database monitoring capabilities by enabling Database
Insights for load testing infrastructure.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added automatic launch agent support for Fleet Desktop.app on macOS
devices, enabling the app to start automatically at system login without
user interaction
* Implemented MDM configuration profiles and policies for seamless
deployment and automatic management across macOS devices
* Added dynamic labeling capabilities to identify and track macOS hosts
with the Fleet Desktop launch agent profile installed
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#34464
# Checklist for submitter
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [x] Timeouts are implemented and retries are limited to avoid infinite
loops
## Testing
- [x] Added/updated automated tests
- [x] QA'd all new/changed functionality manually
---
## What
GitOps `--dry-run` was succeeding when `macos_manual_agent_install` was
set to `true` and a `macos_script` was configured under
`setup_experience`, but the actual GitOps run would fail with:
```
Couldn't add setup experience script. To add script, first disable macos_manual_agent_install.
```
## Why
The `manual_agent_install` conflict validation only existed server-side
in `ee/server/service/setup_experience.go:SetSetupExperienceScript()`.
The script upload call (`uploadMacOSSetupScript()`) was gated by
`!opts.DryRun` in `server/service/client.go`, so during dry-run the
upload was skipped entirely and the validation never fired.
## Fix
Added client-side validation in `server/service/client.go` at the point
where the YAML-parsed `MacOSSetup` struct is processed — before the
script file is validated and loaded. This check runs for **both dry-run
and real runs**, catching the conflict early. Two code paths were fixed:
1. **Team path** (~line 803): Checks `setup.ManualAgentInstall.Value`
when `setup.Script.Value` is set
2. **No-team path** (~line 2603): Checks
`macOSSetup.ManualAgentInstall.Value` when `macOSSetup.Script.Value` is
set
## How I reproduced the issue locally
### Prerequisites
- MySQL and Redis running via Docker: `docker compose up -d mysql_test
redis`
### Steps
1. Wrote an integration test
(`TestDryRunMacOSSetupScriptWithManualAgentInstallConflict`) that:
- Creates a GitOps user and fleetctl config
- Creates a bootstrap package server serving `testdata/signed.pkg`
- Creates a `.sh` script file with `echo "setup script"`
- Creates a **global config** YAML (minimal server settings)
- Creates a **team config** YAML with `macos_manual_agent_install:
true`, `macos_script: <path>`, and `macos_bootstrap_package: <url>`
- Runs `fleetctl gitops --dry-run` and asserts it fails
- Runs `fleetctl gitops` (no dry-run) and asserts it fails
2. Ran the test **before the fix** — confirmed the bug:
```
Dry-run error: <nil> ← BUG: should have failed
Real run error: ...status 422...first disable macos_manual_agent_install
← correctly fails
```
3. Applied the fix and re-ran — **both dry-run and real run now fail**
with the `macos_manual_agent_install` conflict error.
### Test command
```bash
MYSQL_TEST=1 REDIS_TEST=1 go test -v \
-run TestIntegrationsEnterpriseGitops/TestDryRunMacOSSetupScriptWithManualAgentInstallConflict \
./cmd/fleetctl/integrationtest/gitops/... -count=1 -timeout 600s
```
Both sub-tests (team and no-team paths) pass. All related existing tests
continue to pass:
- `TestMacOSSetup`, `TestMacOSSetupScriptWithFleetSecret`,
`TestDeletingNoTeamYAML`, `TestDisallowSoftwareSetupExperience`
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* GitOps dry-run now correctly fails when a macOS setup configuration
combines manual agent installation with a provided setup script,
preventing false-positive dry-run success.
* **Tests**
* Added unit and integration regression tests to verify dry-run and
real-run rejection of conflicting macOS setup configurations for both
team-scoped and unassigned host scopes.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
related to [#44391](https://github.com/fleetdm/fleet/issues/44391)
This will add 10 info level log statements during vulnerability
scanning.
Example:
`ts=2026-05-03T18:32:26Z level=info msg="phase completed"
cron=vulnerabilities phase=nvd elapsed=59.450125s`
## Testing
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Improvements**
* Vulnerability scanning now logs overall elapsed time and per-phase
durations so operators can see how long full scans and each scanner
phase take.
* **Chores**
* Repository ignore settings updated to exclude an additional path
(non-functional housekeeping).
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: ksykulev <230639+ksykulev@users.noreply.github.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
Added MDM components to our high level architecture doc. Is it too much?
I could see it going either way. My argument for it not being too much
is that ultimately it's all very tightly intertwined now and I almost
feel like the big picture isn't complete without it
---------
Co-authored-by: George Karr <georgekarrv@users.noreply.github.com>
- Disable performance insights
- Allow redis instance count >=1
- Properly set ecs_cluster logging config path
- Targeted apply with auto approve for pre-creating fleet and execution
roles
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Enhanced ECS cluster logging with CloudWatch integration and extended
log retention to 365 days.
* Adjusted RDS monitoring configuration and disabled performance
insights for operational optimization.
* Reduced minimum Redis instance requirement from 3 to 1 for greater
deployment flexibility.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Add a new macOS configuration profile
(fleet-desktop-launchagent.mobileconfig) that configures managed
login/background items to allow Fleet Desktop.app's launch agent to run
at login without prompting the user. Also reference the profile in
it-and-security/fleets/workstations.yml so hosts labeled as having Fleet
Desktop.app installed will match the control.
Delete the macOS policy 'enrollment-profile-up-to-date' and the 'Nudge
test devices' label, and remove their references from it-and-security
manifests. Files removed:
it-and-security/lib/macos/policies/enrollment-profile-up-to-date.yml and
it-and-security/lib/all/labels/nudge-test-devices.yml; references
dropped from it-and-security/default.yml,
it-and-security/fleets/testing-and-qa.yml, and
it-and-security/fleets/workstations.yml to avoid loading deleted
artifacts.
Automated ingestion of latest Fleet-maintained app data.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Updated AnyDesk macOS version support to 9.7.0
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Co-authored-by: allenhouchins <32207388+allenhouchins@users.noreply.github.com>
**Related issue:** Resolves#42887.
From Claude's audit:
```
[...]
Concerns worth addressing
A. Catalog drift is the real long-term risk. Today the yaml is curated.
If a future engineer adds (say) POST /users/api_only, PATCH /users/api_only/:id, POST /users/roles/spec,
POST /password_reset, or any session-issuing route, an allowlisted api_only user can clone themselves or
broaden a peer's allowlist.
Suggest a CI test that hard-fails if any of those route prefixes show up in api_endpoints.yml,
plus a comment at the top of the yaml listing the categories that must never be added (user/role/invite/password/session/SSO).
[...]
```
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Tests**
* Added validation tests for API endpoint configuration to ensure
security compliance and proper detection of restricted endpoint
combinations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
**Related issue:** Resolves#44190
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
- [x] Timeouts are implemented and retries are limited to avoid infinite
loops
## Testing
- [x] Added/updated automated tests
- [ ] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added a periodic cleanup job that removes aged, acknowledged Windows
MDM command-queue entries to reduce write pressure during ACK
processing.
* **Bug Fixes**
* Pending-command detection now excludes already-ACKed commands from
dispatch; queue rows are retained after ACK and cleaned later.
* **Tests**
* Added and updated tests to validate cleanup behavior and revised
ACK/queue semantics.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Version bumped to v4.84.2 across Helm charts, container images,
deployment infrastructure, and npm package.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Delete the macOS configuration profile that managed Fleet Desktop as a
login item
(it-and-security/lib/macos/configuration-profiles/fleet-desktop-login-item.mobileconfig)
and remove its inclusion and associated label filter from the
workstations fleet controls (it-and-security/fleets/workstations.yml).
Cleans up an obsolete/unused profile and its reference.
Update fleet-desktop from 1.1.0 to 1.2.0 across Homebrew cask, API
metadata, and darwin output. Adjust package URLs and filenames, update
SHA256 checksums and ruby_source_checksum, and refresh installer script
reference and patched/version queries to match the new release.
Automated ingestion of latest Fleet-maintained app data.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Updated application version metadata for 8x8 Work, Adobe Acrobat
Reader, Cursor, Ollama, Postman, and Zed across supported platforms with
latest release versions and integrity verification checksums.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Co-authored-by: mostlikelee <16102903+mostlikelee@users.noreply.github.com>
Make Fleet Desktop detection more precise and switch to fleet-maintained
app workflow. Rename label to include the .app suffix and filter by
bundle_identifier; update workstations.yml to reference the new label
and replace the file-based software entry with the fleet-maintained slug
(fleet-desktop/darwin). Add a Fleet Desktop entry to the
patch-fleet-maintained-apps policy so Fleet Desktop is managed via the
fleet_maintained_app_slug. Remove the now-obsolete update-fleet-desktop
policy and the packaged fleet-desktop software metadata file.
## Summary
- Enables host expiry settings for three teams by setting
`host_expiry_enabled: true` and `host_expiry_window: 90` (days) in each
team's GitOps configuration file.
- Hosts that have not checked in for 90 days will be automatically
removed from these teams' inventories.
## Changes
- `it-and-security/fleets/workstations.yml`: Updated
`host_expiry_settings` to enable expiry with a 90-day window.
- `it-and-security/fleets/company-owned-mobile-devices.yml`: Updated
`host_expiry_settings` to enable expiry with a 90-day window.
- `it-and-security/fleets/personal-mobile-devices.yml`: Updated
`host_expiry_settings` to enable expiry with a 90-day window.
## Teams affected
| Team | File | Window |
|------|------|--------|
| 💻 Workstations | `workstations.yml` | 90 days |
| 📱🏢 Employee-issued mobile devices | `company-owned-mobile-devices.yml`
| 90 days |
| 📱🔐 Personal mobile devices | `personal-mobile-devices.yml` | 90 days |
## Related
- Follows the same pattern as #44649 (Testing & QA team, 30 days).
---
Built for [Allen
Houchins](https://fleetdm.slack.com/archives/D0AFASNBZMW/p1777903880443949?thread_ts=1776095784.812489&cid=D0AFASNBZMW)
by [Kilo for Slack](https://kilo.ai/slack)
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
## Summary
- Enables host expiry settings for the "🧪 Testing & QA" team by setting
`host_expiry_enabled: true` and `host_expiry_window: 30` (days) in the
team's GitOps configuration file.
- Hosts that have not checked in for 30 days will be automatically
removed from this team's inventory.
## Changes
- `it-and-security/fleets/testing-and-qa.yml`: Updated
`host_expiry_settings` to enable expiry with a 30-day window.
---
Built for [Allen
Houchins](https://fleetdm.slack.com/archives/D0AFASNBZMW/p1777903609156429?thread_ts=1776095784.812489&cid=D0AFASNBZMW)
by [Kilo for Slack](https://kilo.ai/slack)
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
<!-- Add the related story/sub-task/bug number, like Resolves#123, or
remove if NA -->
**Related issue:** Resolves#44533
Adds an option to return a PEM certificate from the request_certificate
endpoint, rather than the PKCS7 envelope an EST server returns. This
allows it to be more easily used in scripts without conversions, at the
(small) cost of among other things dropping the PKCS7 envelope which
could be signed by the server, etc(though the PEM cert itself should
also be)
# Checklist for submitter
If some of the following don't apply, delete the relevant line.
- [x] Changes file added for user-visible changes in `changes/`,
`orbit/changes/` or `ee/fleetd-chrome/changes`.
See [Changes
files](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/guides/committing-changes.md#changes-files)
for more information.
- [x] Input data is properly validated, `SELECT *` is avoided, SQL
injection is prevented (using placeholders for values in statements), JS
inline code is prevented especially for url redirects, and untrusted
data interpolated into shell scripts/commands is validated against shell
metacharacters.
## Testing
- [x] Added/updated automated tests
- [x] Where appropriate, [automated tests simulate multiple hosts and
test for host
isolation](https://github.com/fleetdm/fleet/blob/main/docs/Contributing/reference/patterns-backend.md#unit-testing)
(updates to one hosts's records do not affect another)
- [x] QA'd all new/changed functionality manually
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* The "Request a Certificate" endpoint can optionally return the issued
certificate as a PEM-encoded X.509 CERTIFICATE block instead of a
PEM-encoded PKCS#7 envelope.
* **Tests**
* Added comprehensive tests covering PEM conversion, tolerance for
base64 whitespace/newlines, error handling for malformed PKCS#7, and
multi-certificate envelope cases.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
@noahtalerman: For the following bug:
- https://github.com/fleetdm/fleet/issues/43160
## Summary
- Condense and relocate the S3 `endpoint_url` override warning from the
general `## S3` section header into the individual `endpoint_url`
setting sections
- The primary warning now lives in
`s3_software_installers_endpoint_url`; the `s3_carves_endpoint_url` and
`packaging_s3_endpoint_url` sections cross-reference it
- Condense region-discovery notes similarly (primary in
`s3_software_installers_region`, cross-references elsewhere)
- Condense the Reference Architectures S3 section to a single paragraph
## Changes
**`docs/Configuration/fleet-server-configuration.md`:**
- Removed the large callout block under `## S3`
- Added concise warning to `s3_software_installers_endpoint_url` with
link to [#42349](https://github.com/fleetdm/fleet/issues/42349)
- Added concise note to `s3_software_installers_region`
- Updated `s3_carves_endpoint_url`, `s3_carves_region`,
`packaging_s3_endpoint_url`, and `packaging_s3_region` with short
cross-references
**`docs/Deploy/Reference-Architectures.md`:**
- Condensed the `S3 configuration for AWS deployments` section to a
single paragraph with link to the config reference
## Related
- Relates to #42349
---
Built for
[ntalerman](https://fleetdm.slack.com/archives/D0AEA6U4SM9/p1775051868003859?thread_ts=1775051234.724799&cid=D0AEA6U4SM9)
by [Kilo for Slack](https://kilo.ai/features/slack-integration)
---------
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
Co-authored-by: Noah Talerman <47070608+noahtalerman@users.noreply.github.com>
Found this missed TODO from some old story work.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Enhanced OTA enrollment profile validation to properly enforce
end-user authentication requirements when configured, now returning
appropriate error responses for requests missing required authentication
credentials.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Add macOS and Windows screen-lock configuration profiles and manual
exclusion labels, and wire them into fleet manifests and policies.
- Add macOS mobileconfig (screen-lock-inactivity) to start screensaver
after 900s and require a password with a 60s delay.
- Add Windows configuration (Screen lock timeout.xml) to set
InteractiveLogon_MachineInactivityLimit to 900s (15 minutes).
- Create manual labels: "macOS screen lock exclusions" and "Windows
screen lock exclusions" (empty host lists).
- Register the new labels in it-and-security/default.yml and include the
new profiles in workstations.yml with labels_exclude_any pointing to the
appropriate exclusion label.
- Update macOS and Windows policy YAMLs to exclude hosts in the
corresponding exclusion labels.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added macOS and Windows "screen lock exclusions" labels to allow
manual exemption of devices.
* Introduced a macOS configuration profile that enforces a 15-minute
inactivity screen lock and requires a password on resume.
* Introduced a Windows configuration profile setting an equivalent
15-minute inactivity timeout.
* Screen-lock policies now support label-based exclusions so exempted
devices are not affected.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Automated ingestion of latest Fleet-maintained app data.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Updated macOS application configurations to support new versions:
* DBeaver Community 26.0.4
* Lulu 4.3.2
* Stats 2.12.13
* Warp 0.2026.04.29.08.57.01
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Co-authored-by: mostlikelee <16102903+mostlikelee@users.noreply.github.com>
The current example under verify encryption results in an error, as
ENCRYPTED is an unknown column
```
lsblk -o NAME,MOUNTPOINT,TYPE,SIZE,FSUSED,FSTYPE,ENCRYPTED
lsblk: unknown column: ENCRYPTED
```
- https://fleetdm.slack.com/archives/C09JAPRM1EJ/p1777467446591059
Update the enroll hosts guide to use the 3.14.1 version of the WiX
binaries. Attempting to use the 3.11 version results in an error when
passing the `--arch` flag.
Automated ingestion of latest Fleet-maintained app data.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Chores**
* Updated application version metadata for Adobe Acrobat Reader,
Dialpad, Granola, OneDrive, and TablePlus to reflect the latest releases
across Windows and macOS platforms.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Co-authored-by: mostlikelee <16102903+mostlikelee@users.noreply.github.com>
Pin the gnome-shell appindicator extension to a specific ubuntu upstream
commit and download an immutable tarball URL. Verify the downloaded
archive's SHA-256 (refuse to proceed if sha256sum/shasum is missing or
the checksum mismatches) and only extract when verification succeeds.
Add quoting for $username in sudo/mkdir invocations and preserve the
staging/copy workflow for user-owned extension installation. For the
InstallRemoteExtension flow on other distros, add a 90s polling timeout
when waiting for metadata.json and surface an error on timeout to avoid
hanging indefinitely.
This pull request addresses a startup issue with Fleet Desktop on
openSUSE Leap 16 and similar Linux distributions. The main change is to
adjust how Fleet Desktop and key-escrow dialogs are launched to avoid
environment variable loss caused by login shell profile scripts. The fix
is scoped specifically to openSUSE Leap 16+ to avoid impacting other
distributions.
**Distribution-specific sudo invocation changes:**
* The `-i` (login shell) flag is now omitted from the `sudo` command
when launching Fleet Desktop and key-escrow dialogs on openSUSE Leap 16
and newer, preventing environment variables from being lost due to
profile script interference.
[[1]](diffhunk://#diff-633ab361af6795ef458233819e2806dfba4ca56f684866d956321825b8fd2e91R1)
[[2]](diffhunk://#diff-3e8315d9f12512bce490457c5d20bd7c5aebaa2a8e18b1abf50e504815dd7a9dR178-R193)
* For all other supported distributions, the previous behavior (using
`-i`) is preserved to maintain compatibility and avoid unnecessary
re-testing.
**Detection logic:**
* Introduced a new helper function `isOpenSUSELeap16Plus` in
`execuser_linux.go` to detect if the host is running openSUSE Leap 16 or
newer by parsing `/etc/os-release`. This ensures the workaround is only
applied where necessary.
---
**Related issue:** N/A — surfaced via field investigation on openSUSE
Leap 16 (arm64).
This PR addresses two distinct issues that together prevent Fleet
Desktop from working on openSUSE Leap 16, both validated end-to-end on a
real Leap 16 (arm64) host.
## 1. Launch reliability — drop `sudo -i`
`orbit/pkg/execuser/execuser_linux.go`
On Linux, Orbit launches Fleet Desktop with:
```
sudo -n -i -u <user> -H env WAYLAND_DISPLAY=… … FLEET_DESKTOP_DEVICE_IDENTIFIER_PATH=/opt/orbit/identifier … /…/fleet-desktop
```
The `-i` flag makes sudo "simulate initial login" — it runs the target
user's shell as a login shell and wraps the rest of the command in `bash
--login -c '<escaped>'`. That sources `/etc/profile` and every script in
`/etc/profile.d/*` before our `env KEY=val … fleet-desktop` line runs,
and shell metacharacters (`=`, `:`, `/`, `.`) get backslash-escaped
through the shell layer.
On **openSUSE Leap 16 (arm64)**, that indirection causes the inline
env-var assignments to not reach `fleet-desktop`, which exits
immediately with:
```
FTL missing URL environment FLEET_DESKTOP_DEVICE_IDENTIFIER_PATH
```
Orbit then respawns it every ~15 s in a tight kill-and-respawn loop, so
the tray icon never appears.
**Fix:** drop `-i` from the sudo invocation. We don't need a login
shell:
- `-H` already sets `HOME` to the target user.
- sudo's default `env_reset` sets `USER` / `LOGNAME` / `SHELL` / `MAIL`
and `PATH` to `secure_path`.
- All session vars (`WAYLAND_DISPLAY`, `DISPLAY`,
`DBUS_SESSION_BUS_ADDRESS`, `LD_LIBRARY_PATH`) and every
`FLEET_DESKTOP_*` var are already passed explicitly via `env KEY=val …`.
After the change, sudo `execve()`s `env` directly with no shell layer in
between, so `/etc/profile.d` sourcing and shell-escaping are out of the
picture.
The `runuser -l` /proc/keys-leak regression from PR #32309 does not
apply — that was specific to `runuser -l` creating session keyrings;
sudo without `-i` doesn't.
# Checklist for submitter
- [x] Changes file added:
`orbit/changes/fleet-desktop-linux-no-login-shell`
- [x] Input data is properly validated; untrusted data interpolated into
shell scripts/commands is validated against shell metacharacters.
- [x] Timeouts are implemented and retries are limited to avoid infinite
loops (script's wait loop now bounded at 90s).
- [x] If paths of existing endpoints are modified without backwards
compatibility, checked the frontend/CLI for any necessary changes — N/A.
## Testing
Manual QA needed before merge:
- [x] **openSUSE Leap 16 (arm64)** — Fleet Desktop process starts, stays
running, env vars present, no FTL respawn loop. Done via `sudo` shim.
- [x] **openSUSE Leap 16 (arm64) — extension fallback** — manual tarball
install + schema compilation produces a working tray icon (matching what
the script automates).
- [ ] **Ubuntu 22.04 / 24.04** — regression check: Fleet Desktop tray
icon still appears, key-escrow zenity dialog still renders, AppIndicator
script still installs via the official path.
- [ ] **Fedora (recent)** — regression check: same as above.
- [ ] **Debian** — regression check: same as above.
- [ ] **openSUSE Tumbleweed** — confirm `InstallRemoteExtension` path
still works (no fallback path triggered).
## fleetd/orbit/Fleet Desktop
- [x] Verified compatibility with the latest released version of Fleet —
pure launch-flag change plus a script update; no protocol or schema
impact.
- [x] If the change applies to only one platform, confirmed that
`runtime.GOOS` is used as needed to isolate changes — Go change is in
`execuser_linux.go`, only built on Linux. The script is Linux-only by
construction.
- [ ] Verified that fleetd runs on macOS, Linux and Windows — Linux
re-verification pending QA above; macOS/Windows code paths unchanged.
- [ ] Verified auto-update works from the released version of component
to the new version.
## Notes for reviewers
- The tray-icon visibility issue is an OS-side prerequisite (GNOME 3.26+
has no native tray), so the AppIndicator extension is required
regardless. Even after installing it, Wayland requires a logout/login to
pick up new extensions — this is documented behavior and not specific to
the fallback path.
Noah Talerman