Update subsite upstreams and runtime references together with the parent site. Preserve stopped sites and custom PHP settings, and restore configurations and database records when applying the change fails.
Version official app package and compose downloads using the synchronized store revision. Refresh outdated local resource caches and use synchronized database parameters for official PHP version options.
* feat(container): support local image uploads and improve import task logs
* feat(container): support local image uploads and improve import task logs
* fix(container): remove duplicate image load response helper
* fix(dashboard): keep disk info working when a mount is stale or broken
The dashboard, the file manager mount list and the alert disk list
enumerated mounts with df. df calls statfs on every mount, so one bad
mount took the whole disk list down: the request hung until the
frontend timed out, or the list came back empty.
- enumerate mounts from /proc/self/mountinfo instead of running df
- read each mount through a guard with a 3s timeout that keeps at most
one call per mount in flight
- fail closed: a mount that cannot be read stays in the list with
errorType and error, and an unreadable mount table is an error
rather than an empty list
- show failed mounts and their reasons on the dashboard and in the
file manager
- cmd: after killing a timed-out command, wait at most 5s for it
instead of forever
* fix(dashboard): distinguish remounted and hidden filesystems
* fix(disk): simplify disk queries and bound timeout handling
* fix(disk): prevent overlapping df collection tasks
---------
Co-authored-by: ssongliu <sloooop1x@gmail.com>
Replace database rule synchronization with native rule operations and file-based backup, import, and recovery. Update firewall management views, descriptions, whitelist handling, and translations.
Include terminal reconnect fixes and xpack build tags for Darwin build targets. Remove temporary regression test files before committing.
The SFTP uploader used io.Copy(dstFile, srcFile). Since the source is an
*os.File it implements io.WriterTo, so io.Copy took the os.File.WriteTo
branch and issued one synchronous SFTP WRITE per packet, waiting for the
server reply every time. Throughput was therefore bounded by packetSize/RTT
instead of the link capacity: on a ~68ms RTT link a 440MB backup reached
only ~0.25 MB/s, while rclone / OpenSSH sftp reach 16-19 MB/s on the very
same link and target.
pkg/sftp already provides a pipelined write path (File.ReadFrom, enabled by
UseConcurrentWrites and bounded by MaxConcurrentRequestsPerFile, default 64).
Use it for uploads, and truncate the remote file to the number of bytes
actually written when the concurrent write fails, to avoid leaving holes.
Measured against the same 68ms RTT target: a 96MiB upload went from 6m40s
(~0.25 MB/s) to ~6s (~16 MB/s); a 950MB three-site backup job from ~2h40m to
~2m04s.
Co-authored-by: 4kercc <13767585+4kercc@users.noreply.github.com>
CityFun
Yuki Guo
Woolsen
Octopusandocto-patch
ssongliu
HynoRandCursor
蘭
CN_CoreStebandssongliu
4kerccand4kercc