Recyclarr v8.0 moved local includes from configs/ to a dedicated
includes/ directory. The install script now creates both dirs
upfront. The update script migrates existing include subdirs
from configs/ to includes/ after the binary update and also
handles detection for both old (recyclarr.yml) and new
(configs/) config layouts.
See https://recyclarr.dev/guide/upgrade-guide/v8.0
Ref #12109
Planka v2.0.1 moved all uploads to a unified data directory:
- public/{favicons,user-avatars,background-images} -> data/protected/
- private/attachments -> data/private/
The update script was still restoring files to the old locations,
which made attachments, avatars and backgrounds disappear after
updating from rc.4 to 2.0.1.
Changes:
- update_script: backup handles both old and new layouts
- update_script: restore now uses new v2 paths
- install script: create v2 data directories on fresh install
Ref #12066
The JVM heap options were appended with >> which left the old
defaults in place. Use sed to replace them instead.
Also add the missing systemctl daemon-reload that ES requires
after package installation before the service can be started.
Closes#12111
* fix(huntarr): add build-essential for native pip dependencies
Some Python packages in requirements.txt need a C++ compiler for
native extensions. Without build-essential the uv pip install
step fails with 'command c++ not found'.
Closes#12117
* Change apt-get to apt for installing dependencies
* Ensure git dependency before updating Huntarr
Added dependency check for git before fetching and deploying the GitHub release.
* drunk
* Add calibre-web (ct)
* Fix casing in Calibre-Web references
* Correct case for Calibre-Web in installation script
---------
Co-authored-by: push-app-to-main[bot] <203845782+push-app-to-main[bot]@users.noreply.github.com>
Co-authored-by: Chris <punk.sand7393@fastmail.com>
* Fix: Add disk space checking for OPNsense VM FreeBSD image decompression
- Add check_disk_space() function to verify available storage
- Check for 20GB before download and 15GB before decompression
- Provide clear error messages showing available vs required space
- Add proper error handling for unxz decompression failures
- Clean up compressed .xz file after decompression to save space
- Add progress messages for download and decompression steps
Fixes issue where script fails at line 611 with 'No space left on device'
when /tmp directory lacks sufficient space for ~10-15GB decompressed image.
* Increase OPNsense VM disk size from 10GB to 20GB
- Provides more space for system updates, logs, and package installations
- 20GB is a more appropriate size for OPNsense production use
Introduce post_progress_to_api() in misc/api.func — a non-blocking, fire-and-forget curl ping (gated by DIAGNOSTICS and RANDOM_UUID) that updates telemetry status to "configuring". Wire this progress ping into multiple scripts (alpine-install.func, install.func, build.func, core.func) at key milestones (container start, network ready, customization, creation, cleanup) and replace/deduplicate some earlier post_to_api calls. Also update error_handler.func to always report failures immediately via post_update_to_api to ensure failures are captured even before/after container lifecycle.
- This PR will add the ability to change the PatchMon listen port in the
Nginx config during update, if the `SERVER_PORT` env var is set in
`/opt/patchmon/backend.env` and if the port is not 443
- If not set, or if set and the port is 443, then no changes are made to
the listen port in the Nginx config
Prevent hangs when pulling logs from containers by wrapping pct pull calls with timeout (8s) and running ensure_log_on_host under timeout (10s). Always send telemetry (post_update_to_api) before attempting best-effort log collection so status is reported even if log retrieval blocks. Update EXIT/ERR/SIGHUP/SIGINT/SIGTERM traps and consolidate error/interrupt handlers to use the new timeouted log collection. Changes in misc/build.func and misc/error_handler.func.
