The remaining routes of the idle-in-transaction class: the file-manager,
storage, camera-snapshot and timelapse routes each took their printer row
via Depends(get_db) and then talked FTP/camera on the same held session, so
a farm dashboard polling cover/snapshot tiles (offline printers included)
crept the pool to exhaustion over ~23h. They now read in a short session and
release before the I/O; timelapse re-opens a fresh session only for the write.
Also caps the four bare-executor FTP helpers with asyncio.wait_for so a
saturated 48-worker pool can't pin a caller (and its DB connection)
indefinitely, and runs the synchronous smtplib send off the event loop with
an explicit timeout so a wedged relay can't freeze the loop.
The reporter's 19-printer farm started prints "one by one", up to an hour apart.
check_queue awaited each dispatch inline, and a dispatch includes the FTP upload,
so every printer queued behind every other printer's transfer despite being an
independent machine. His logs give the arithmetic: 40978500 bytes in 254.1s,
157 KB/s - a Bambu printer's SD write, not the network, is the bottleneck. Nineteen
of those in series is ~80 minutes, and the next upload started 131 ms after the
previous one finished. The delay is linear in fleet size, which is why it got worse
the more printers he selected.
Dispatch is now collected during the (still sequential) selection loop and run
concurrently afterwards, capped by queue_max_concurrent_uploads - Settings ->
Workflow -> Queue & Dispatch, default 4, 1 restores the old behaviour. Every gate
is untouched; only the transfers overlap. The pass still awaits its uploads before
returning: _start_print flips the row pending -> printing only after the upload,
so an early return would let the next tick re-dispatch the same rows.
FTP work moves to its own thread pool. It was on asyncio's default executor -
min(32, cpu+4), six threads on a 2-core NAS, shared with everything else - which
was survivable only while uploads were serial.
Two problems the same bundle exposed:
A printer that accepts project_file but never starts (#1678) was retried forever:
270s watchdog, revert to pending, re-upload the whole file, repeat. Hence his
"printer who, since the morning, still not launch" - and on a farm each lap also
eats an upload slot the other printers are waiting on. Attempts are now counted on
the queue item; after three it fails with a message pointing at the printer instead
of queueing a fourth re-upload.
The debug bundle we asked him for held 4m49s of history. The push_status dumps fired
on every frame rather than on change - several while their own comment claimed
otherwise - which is 27,727 of the bundle's 29,830 lines and rolls 5 MB in under five
minutes on 19 printers. They now log transitions only. The bundle also read just the
live log while three rotated backups sat next to it, under a byte budget four times
larger than the file it was reading.
Migration verified on SQLite and Postgres: idempotent, backfills legacy NULLs
(dispatch_attempts + 1 is NULL for a NULL row, which would silently disable the cap).
Tests: 6 on concurrent dispatch (overlap, cap honoured, 1 == serial, default applies
with no settings row, a failed printer does not cancel its siblings, no early return),
4 on the retry budget, 6 on the bundle's rotated-log span, 7 on the debug gating.
Each verified to fail against the unfixed code - the first end-to-end log assertion I
wrote passed without the fix and had to be tightened.
upload_file_async carried a flat 600s wall-clock deadline and ran the
transfer via asyncio.wait_for(run_in_executor(...)). wait_for cancels the
future, not the executor thread. A 96 MB 3MF to an A1 over WiFi sustains
~75 KB/s and needs ~20 minutes, so the await gave up at ~70 MB, returned
False, and with_ftp_retry started a second STOR of the same file onto the
same printer while the first was still streaming. The reporter filmed two
transfers of one job climbing in parallel at 2% and 72%; the print never
landed and the printer read as having a flaky network.
The deadline is now derived from the file size against a 25 KB/s floor, so a
slow-but-healthy transfer can finish — a link that has actually died is
caught within socket_timeout by the blocking sendall, which is what should be
detecting failure. A deadline expiry now stops the transfer for real: the
worker is signalled, raises UploadCancelled from its progress callback, and
upload_file's existing cancel path breaks the send loop and deletes the
partial file. with_ftp_retry never retries that, and a per-printer lock makes
overlapping uploads impossible however they were triggered.
ssl.create_default_context() leaves minimum_version at MINIMUM_SUPPORTED,
so the floor came from the OpenSSL build rather than from Bambuddy. On
identical OpenSSL 3.5.6, python:3.13-slim-trixie reports TLSv1_2 while a
bare-metal venv reports MINIMUM_SUPPORTED -- Docker installs were floored
at 1.2, bare-metal and appliance installs were not.
Set minimum_version explicitly in ImplicitFTP_TLS and the MQTT client. On
the P2S/X2D profiles that also cap maximum_version this becomes an exact
TLS 1.2 pin. Probed against an X1C and an H2D on :990 and :8883: both
complete only on TLS 1.2 and reject 1.0, 1.1 and 1.3; live FTPS login
through the new path succeeds on both.
Also correct a stale comment in ftp_profiles.py -- X1C and H2D refuse
TLS 1.3, so cap_tls_v1_2 is a no-op there, contrary to what it claimed.
Two warnings polluting every A1 support bundle on healthy prints, both
unrelated to the timelapse-default behaviour the issue actually reports.
1. mqtt_bridge.py's post-bind nudge calls request_status_update on the
real printer's MQTT client to populate the bridge cache without
waiting for the next periodic pushall. The bind frequently races the
TLS handshake, especially on A1 firmware. Skip the nudge when
state.connected is False — the periodic pushall fills the cache
anyway. The WARNING in bambu_mqtt.py stays for the genuinely-
actionable callers (refresh-status API, bug reporter).
2. Post-finish SD-card cleanup (and the symmetric forced-timelapse dir
walk) used delete_file_async's bool return to drive a WARNING when
all candidates failed. A1 firmware self-cleans the SD card before
our cleanup runs — every candidate FTP-DELE returns 550, we burn
the retry budget, then WARN on a successful print. Introduce
DeleteResult.{DELETED,NOT_FOUND,FAILED} so the helpers only WARN
on real network/auth/transient failures. NOT_FOUND advances to the
next candidate without consuming the 2s backoff. User-facing delete
endpoint returns 404 on NOT_FOUND.
Python 3.13 negotiates TLS 1.3 by default. The P2S firmware 01.02.00.00
vsFTPd build doesn't tolerate TLS 1.3's async session-ticket model on
the FTPS data channel — session resumption races, the data channel gets
torn down mid-stream, uploads land truncated at a chunk boundary, and
the printer replies 426 instead of 226. Visible to the user as "unable
to parse 3mf file" 30 s into the print.
Capping the SSL context's maximum_version to TLS 1.2 makes session
resumption synchronous and uploads complete normally.
Follow the per-model pattern established by camera_profiles.py in the
#1395 follow-up: add backend/app/services/ftp_profiles.py with a frozen
FTPProfile dataclass and a per-model registry. Only P2S (display name
+ N7 SSDP code) gets the cap today. X1C, H2D, P1S, A1 stay on negotiated
TLS 1.3 — the maintainer's dogfooded printers see zero behaviour change.
Previous daily build (1fac0276) tightened the post-STOR voidresp
handler to fail on any ftplib.Error, stopping Bambuddy from
sending a print command for a truncated 3MF. Reporter
(@enjoylifenow on a P2S) then confirmed — after a clean SD-card
filesystem check, reformat, and power cycle — that v0.2.4.1
worked on the same hardware. That proves the 426 returned by
this firmware revision is noise: the TLS data-channel close
races the 226 confirmation, server reports failure, file is in
fact on the SD card.
Reverting wholesale would re-introduce the silent-truncation
bug from the original fix. Narrow the rule instead: after an
ftplib.Error from voidresp, run an FTP SIZE against the upload
path. SIZE matches the local file size → warn and proceed
(the reporter's case). SIZE mismatch, or SIZE itself raises →
fail loudly with full diagnostics (the original tightened
behavior — preserved).
Applied identically to upload_file() and upload_bytes() so the
A1-compatibility manual-transfer path is covered.
Tests: two regressions from the previous round renamed and
split into intact / truncated / size-check-fails. Intact-file
tests inject SIZE explicitly because pyftpdlib only flushes on
a clean voidresp — which can't happen when we monkeypatch
voidresp to raise. Docstring spells that out. 87 FTP unit tests
green; 118 FTP-touching tests across unit+integration green;
ruff clean.
The View-Timelapse-greyed-out behavior #1417 was originally
about stays untouched; once the reporter confirms upload
reliability is back, that diagnosis continues on a healthy
install.
bambu_ftp.upload_file (and upload_bytes) wrapped the voidresp() call in a
broad "except Exception: log warning and proceed" because H2D printers
can take 30+ seconds to send the 226 and we don't want to fail on that.
But the same handler was swallowing ftplib.error_temp (e.g. 426 "Failure
reading network stream") from buggy printer firmware, which explicitly
means the data stream was cut mid-transfer and the file on the SD card
is partial.
Bambuddy then sent the print command anyway, and the printer surfaced a
generic "unable to parse 3mf file" error 30 seconds into the print
attempt -- with nothing in the log on the user side to suggest the
upload had actually failed.
Split the catch: ftplib.Error subclasses (server-reported failure)
re-raise so the outer handler returns False; everything else (socket
timeout etc.) keeps the existing proceed-with-warning behaviour so the
H2D 226 tolerance survives.
Two regression tests patch _ftp.voidresp to raise error_temp("426 ...")
and assert both upload_file() and upload_bytes() return False.
The underlying P2S firmware / TLS-data-channel issue that triggers the
426 for the reporter is separate -- this change just stops Bambuddy from
hiding it.
Daily builds since 889c8bd8 (Apr 29) silently destroyed archive copies
and library file bytes on every print completion. Reprint / View G-code
later returned 404 with no log line explaining why; the DB row was
intact and the archive grid kept showing the entry, but the file
behind archive.file_path no longer existed on disk.
Root cause: #1166 added three dispatch sites that cache the live
archive copy (and library file bytes for Direct-Print) in the shared
3MF download cache, so /cover could skip a redundant FTP transfer
mid-print. The cache was originally designed for transient downloads
under archive_dir/temp/, and clear_3mf_cache(printer_id) — called
from on_print_complete to keep that temp dir from accumulating —
happily unlink()'d every cached path. Path.exists() guarded the
unlink, so no exception, no warning, just silent destruction. Listing
didn't change; only acting on the archive surfaced the 404.
Fix: clear_3mf_cache._maybe_unlink refuses to unlink any path outside
archive_dir/temp. Cache dict is still cleared (so re-cache continues
to work and /cover hits a fresh path next print), only the on-disk
delete is gated. Persistent locations — archive/<printer_id>/...,
archive/unassigned/... (VP-archived prints with printer_id=None),
library_files/..., is_external library mounts — all survive.
Regression test test_clear_does_not_delete_persistent_files pins the
contract end-to-end: archive 3mf, library 3mf, and temp 3mf all
cached for the same printer; after clear, all three cache entries
are dropped from the dict, but only the temp file is unlinked from
disk. Two existing tests updated to put fixtures under
archive_dir/temp.
Second wave of #972 — reproducer on a 37.5 MB BambuStudio print to an A1
showed three stacking root causes when Bambuddy restarts mid-print.
1. Archive start_time lost on container restart. The name-based dedup
cancelled any "printing" archive older than 4h and recreated it with
started_at=now(), so a 13h print that saw a restart 10h in ended up
showing ~1.5h duration. Persist MQTT subtask_id on every archive and
match on that first, regardless of age — same id means same print,
resume in place. Also revives Stale-cancelled rows for users
upgrading mid-print.
2. 3MF FTP search tried non-existent paths for ~48 min. Order was
/cache → /model → /data → /data/Metadata → / with 11×30s retries
each; BambuStudio actually pushes to / on A1, so the real path was
tested last. Reorder to / first, and raise a new FileNotOnPrinterError
sentinel from download_to_file on 550 so with_ftp_retry short-circuits
via non_retry_exceptions. 425 / SSL EOF / connection resets still
retry as before.
3. Cover endpoint and archive flow downloaded the same 36 MB twice and
competed for the printer's single FTP socket, producing 425 errors
that fed cause-2's retry storm. Add an in-memory _threemf_path_cache
keyed on (printer_id, normalized filename); whichever flow fetches
first populates it, the other reuses the file read-only. Eviction
runs on on_print_complete and deletes the temp file.
Backend: 14 new tests across test_bambu_ftp.py and a new
test_subtask_archive_resume.py. Existing suite: 2737 pass. ruff clean,
frontend build clean.
Two-part root cause for missing photos/filament/cost on large prints
(#972). The configured ftp_timeout was only plumbed through as the FTP
socket timeout; the asyncio.wait_for wrapping run_in_executor stayed on
its 60s hardcoded default, so the user's 300s setting never applied.
Worse, asyncio.wait_for cannot cancel run_in_executor threads — after
the 60s outer timeout fired, the executor thread kept running
ftplib.retrbinary and frequently completed the download ~30–60s later,
but by then the async wrapper had returned False. with_ftp_retry kept
re-attempting the same path, each retry truncating the file the zombie
thread had just written, and the archive was ultimately persisted as a
fallback with no 3MF.
download_file_async now accepts timeout at each call site (plumbed from
ftp_timeout) and salvages post-timeout success via an explicit
completion flag the executor thread sets only after download_to_file
returns True. Per-attempt completion dict so a prot_p zombie can't
flip the flag for a later prot_c attempt. A cosmetic // prefix in the
directory-search download path is also fixed by replacing string
concatenation with posixpath.join.
The FTP upload skipped reading the server's "Transfer complete"
confirmation, sending the print command before the file was flushed
to the SD card. Now waits for the 226 response with a 60s timeout.
FTP upload: reduce chunk size from 1MB to 64KB for smooth progress bar
updates (~1s intervals instead of 20+ second gaps). Skip voidresp() for
all printer models — H2D delays the 226 response by 30+ seconds after
data transfer, causing a hang at 100%. Add transfer speed and TLS
handshake timing to logs for diagnosing slow connections.
Print/Schedule modal: widen from max-w-lg (512px) to max-w-2xl (672px)
to accommodate long filament profile names like "PLA Support for PETG
PETG Basic @Bambu Lab H2D 0.4 nozzle".
This allows users sending particularly large prints to slow printers such as the P1-series to start a print and then move onto another task in Bambuddy immediately (such as starting more prints on more printers). It also gives the user visibility into what's happening instead of a loading indicator appearing for an indefinite period of time.
The new toast-based UI uses websockets to update in real time. It will also appear for other users / instances of Bambuddy, not just the user who started the prints, allowing more transparency and handling cases where the user closes the page and then comes back wanting to know the status of the dispatching.
fix: review fixes for background dispatch PR #408
- Restore missing imports in main.py (inventory, print_log, virtual_printers, mqtt_smart_plug_service)
- Guard voidresp() for A1 printers to prevent hang after upload
- Don't fail upload on voidresp() error since data transfer already completed
- Add ams_mapping to register_expected_print calls for Spoolman usage tracking
- Fix cancel_job TOCTOU race by using single lock acquisition
- Fix batch counter reset TOCTOU by re-checking condition inside second lock
- Add backgroundDispatch translations to fr.ts and pt-BR.ts
- Remove dead upload_progress_callback definitions
- Skip redundant "Print queued" toast in reprint mode (dispatch toast handles it)
- 5 backend tests: cancel_job single-lock TOCTOU, batch reset re-check,
job lifecycle
- 2 FTP regression tests: voidresp error handling (upload-loop fix),
A1 model voidresp skip
- 1 frontend test: reprint toast suppression
- CHANGELOG: background dispatch feature + test coverage entries
- README: add background dispatch to Scheduling & Automation
- Website: add feature item to Automation section
- Wiki: add Background Print Dispatch section to print-queue.md
Dual-nozzle H2D/H2D Pro: filament matching now respects nozzle assignments
from the 3MF file. Each AMS unit feeds a specific nozzle (L/R), and the
scheduler/frontend constrain matching to only trays on the correct nozzle.
Falls back to unfiltered matching when no trays exist on the target nozzle.
L/R badges shown in the filament mapping UI. Translated in en/de/ja/it.
Fix AMS slot config overwritten on startup: on_ams_change unconditionally
unlinked BL spool assignments on every MQTT pushall, then re-assigned them
sending ams_filament_setting without setting_id — clearing the printer's
filament preset. Now compares spool RFID identifiers before unlinking.
Fix BL spool detection false positives: removed tray_info_idx from detection
logic in both backend is_bambu_lab_spool() and frontend isBambuLabSpool().
Third-party spools using Bambu generic presets had GF-prefixed tray_info_idx
values, causing misidentification. Now uses only tray_uuid and tag_uid.
SQLite WAL mode with 5s busy timeout reduces "database is locked" errors.
download_to_file() returned True when retrbinary transferred 0 bytes
without raising an exception. This caused the /cover endpoint to hit
the empty file check and raise HTTP 500 instead of retrying or
returning 404.
Now treats 0-byte downloads as failures, allowing the cover endpoint's
retry logic to work and falling back to 404 if all attempts fail.
The CodeQL cleanup in "Housekeeping" (2b11efd) bulk-narrowed except
clauses across 50+ files, breaking FTP uploads on ALL printer models.
ftplib.error_perm (550 errors) is not a subclass of ftplib.error_reply,
so diagnose_storage() CWD failures escaped the handler and prevented
STOR from ever executing — causing 100% upload failure and HTTP 500s
on /api/v1/archives/{id}/reprint and /api/v1/library/files/{id}/print.
FTP fixes:
- Remove diagnose_storage() from upload hot path
- Change all except (OSError, ftplib.error_reply) to
except (OSError, ftplib.Error) across bambu_ftp.py
Exception handling reverts (9 files):
- Revert narrowed except clauses back to except Exception in route
handlers and service code where broad catches are intentional
defensive programming (archive parsing, HTTP clients, 3MF/ZIP
processing, Home Assistant, firmware checks)
- Keep narrow exceptions only where safe (single-op blocks like
int(), file.unlink(), socket.close())
- Remove unused XMLParseError imports from archive.py, threemf_tools.py
Closes#287
The A1 printer's FTP server hangs when Python's storbinary() calls
voidresp() to wait for the server's completion response. This caused
upload timeouts on A1 and A1 Mini printers.
Fix contributed by an A1 user - replaces storbinary() with manual
chunked transfer using transfercmd() + sendall():
- Uses 1MB chunks (CHUNK_SIZE constant) for better throughput
- Sets explicit 120s socket timeout on data connection
- Manually closes connection after transfer, avoiding voidresp() hang
Applied to all printer models since the manual approach is compatible
with X1C/P1S/P1P as well (transfercmd is what storbinary uses internally).
User feedback indicated A1 Mini with current firmware works with prot_p
(protected/SSL data channel), not prot_c as previously assumed. Different
A1 firmware versions have different FTP SSL behavior.
Changes:
- Remove hardcoded assumption that A1 models need prot_c
- Try prot_p first for all models (including A1/A1 Mini)
- If upload/download fails on A1 models, automatically retry with prot_c
- Cache working mode per printer IP for subsequent operations
- Add force_prot_c parameter for explicit mode control
This makes FTP work across A1 firmware versions:
- New firmware: prot_p succeeds, cached
- Old firmware: prot_p fails → prot_c fallback succeeds, cached
The FTP code called prot_p() (protected data channel) for all printers,
but for A1/A1 Mini it didn't wrap the data connection in SSL. This
mismatch caused an immediate EOFError - server expected encrypted data
but received plain data.
Fix:
- Use prot_c() (clear/unencrypted data channel) for A1/A1 Mini
- Use prot_p() (protected/encrypted data channel) for X1C/P1S/etc
- Removed non-functional ftplib._SSLSocket = None workaround
A1/A1 Mini: control channel encrypted (implicit TLS), data channel clear
X1C/P1S/etc: both channels encrypted with SSL session reuse
Closes#271
Removed P1S and P1P from SKIP_SESSION_REUSE_MODELS
- These printers use vsFTPd which requires SSL session reuse on data channel
- Only A1/A1 Mini should skip session reuse (they have issues with SSL on data channel)
- P1S/P1P were incorrectly added in commit 9969005, causing EOFError on FTP upload
Closes#266
The fix for A1/P1S FTP uploads (commit 82a6025) was accidentally broken in
commit 9969005 which removed the skip_session_reuse parameter from the
ImplicitFTP_TLS constructor. This caused P2S (and other models in
SKIP_SESSION_REUSE_MODELS) to still use SSL on the data channel, resulting
in "426 Failure reading network stream" errors.
The fix was implemented in commit b96ecfa on test/issue_174 branch but
never merged to main. This cherry-picks that fix.
Also includes:
- Storage diagnostics for debugging upload issues
- Better FTP error logging with specific error codes (553, 550, 552)
- Improved error messages in print scheduler
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Printers without an SD card store files in the root folder `/` instead of
`/cache`. Added root folder to search paths in both main.py and bambu_ftp.py
so 3MF files can be found regardless of storage configuration.
Closes#146
P2S printer creates fallback archives without thumbnails/metadata because
FTP couldn't find 3MF files. The P2S uses different directory structure
(/data/Metadata/) and doesn't have a /cache directory.
Changes:
- Add P2S to SKIP_SESSION_REUSE_MODELS for SSL compatibility
- Add /data/ and /data/Metadata/ to 3MF download paths
- Update fallback search to try multiple directories instead of just /cache
- Add /data paths to FTP storage scan directories
Closes#146
The previous fix only skipped SSL session reuse for A1 printers but
still wrapped the data channel in SSL. VREmma's testing (issue #31)
showed that A1 printers need the data channel to not be SSL-wrapped
at all - they timeout waiting for transfer completion responses.
The fix:
- X1C/P1S: SSL with session reuse on data channel (required by vsFTPd)
- A1/A1 Mini: No SSL on data channel (control channel remains encrypted)
This matches VREmma's working fix (c723f25) which set ftplib._SSLSocket
to None, but implemented properly within the ImplicitFTP_TLS class.
Fixes#31
Thanks VREmma!
Resolves FTP transfer failures on A1/A1 Mini printers caused by SSL
session reuse incompatibility. Also adds configurable FTP timeout
setting for slow WiFi connections.
Changes:
- Add SSL session reuse bypass for A1/A1 Mini printers (auto-detected)
- Add printer_model parameter to FTP functions for model detection
- Add configurable FTP timeout (10-120s, default 30s) in Settings
- Update all FTP callers: archiving, reprint, timelapse, firmware upload
Technical details:
- A1/A1 Mini printers don't support vsFTPd's SSL session reuse
- X1C/P1S continue to use session reuse as required by their FTP server
- ImplicitFTP_TLS now accepts skip_session_reuse flag
- BambuFTPClient auto-detects A1/A1 Mini and sets flag accordingly
Enables checking and uploading firmware updates for printers operating
in LAN-only mode without Bambu Cloud connectivity.
Features:
- Automatic firmware version checking against Bambu Lab servers
- Orange "Update" badge on printer cards when updates available
- Firmware update modal with version info and release notes
- One-click firmware upload to printer SD card via FTP
- Real-time upload progress (actual bytes transferred)
- Step-by-step instructions for triggering update from printer
- Local firmware caching for faster re-uploads
- Supports all Bambu Lab printer models
New files:
- backend/app/services/firmware_check.py - Version checking service
- backend/app/services/firmware_update.py - Upload orchestration
- backend/app/api/routes/firmware.py - REST API endpoints
Also includes:
- FTP upload progress callback support
- 10-minute upload timeout protection
- Firmware cache directory in .gitignore
- AMS "Unexpected format" warnings fixed - P1S/P1P send partial updates without full tray list (per https://github.com/Doridian/OpenBambuAPI/blob/main/mqtt.md). Now handled correctly.
- FTP improvements:
- Added timeouts (60s download, 30s listing)
- Better logging at INFO level
- Shows what files exist in /cache
- Fallback archive - When 3MF can't be downloaded, creates minimal archive so prints are tracked
maziggy