Conversation
…README and .gitignore
The upstream workflow required DockerHub secrets that aren't set on this
fork, blocking builds. Replace with standalone workflows that push
multi-arch images to ghcr.io/${{ github.repository }} using only the
built-in GITHUB_TOKEN, so builds succeed without external credentials.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Collaborator
|
Thanks for your work. Going to leave open for now until I am ready to release the Nfrastack variant of this image. |
…tion and user prompts
latest-* backup files have no extension, causing the compression detection case-statement to fall through and leave decompress_cmd unset. Plain cat then piped raw compressed bytes into jq/the DB client, producing invalid UTF-8 / parse errors and a zero-document restore. Fix: resolve symlinks before extension sniffing, and fall back to `file --brief --mime-type` magic-byte detection when no recognizable extension is found. Applies to all DB types (MySQL, Postgres, MongoDB, Neo4j, CouchDB, Redis). Also adds rdbtools/python-lzf to the image for Redis RDB restore support. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
…d improve Redis restore process
…o non-contiguous configured DB variables
tiredofit
force-pushed
the
main
branch
2 times, most recently
from
August 24, 2026 16:50
51cee5f to
c2fccc6
Compare
backup_couch() dumped each database with a single unbounded GET /$db/_all_docs?include_docs=true request. On a large database (seen in production on an Obsidian LiveSync vault) this ran long enough for CouchDB's own request/memory watermark timeout to abort the response mid-stream, corrupting the JSON and silently failing the nightly backup while the container still exited 0. Replace the single request with a paginated dump using CouchDB's _changes feed (cursor = since, stop condition = pending==0 or a short page), so no single HTTP request/response has to carry the whole database. Output format, filename convention, compression, encryption, checksum, retention cleanup, and failure notification are all unchanged. Page size is tunable via COUCH_BACKUP_PAGE_SIZE (default 2000, global — applies to every DB<NN>_TYPE=couch job). Ports the same logic already verified in production via a runtime bind-mount override on panel-us-e (22,722 docs / 12 pages, ~3m10s vs. a 9m40s timeout on the old approach) into the actual image source. Verified locally against a seeded CouchDB 3 instance (3500 docs / 7 pages at COUCH_BACKUP_PAGE_SIZE=500, output doc count matches live doc_count after excluding deletes) and confirmed the failure path (404/bad credentials) still trips a non-zero exit_code so check_exit_code's error log and failure notification still fire. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
tiredofit
force-pushed
the
main
branch
2 times, most recently
from
September 23, 2026 20:19
b9505ad to
1d93939
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds backup and restore support for Neo4j and CouchDB, rebased onto the latest
main(4.1.100). This supersedes #448 which only covered Neo4j.Neo4j
cypher-shellusingapoc.export.cypher.all()→.cypherfilescypher-shellquerycypher-shellinstalled from Neo4j community edition (NEO4J_VERSION=5.26.0)CouchDB
The project previously had skeletal CouchDB support (type detection + availability check), but
backup_couch()was missing the actual backup command, and there was no restore support.curl+_all_docs?include_docs=true) — no client binary neededDB_NAME=ALLto enumerate all non-system databases via_all_dbsDB_NAMElistDB_NAME_EXCLUDEwhen usingALL_) filtered automatically.jsonextension)_revfrom documents viajq, posts via_bulk_docsbackup_job_backup_job_db_port→backup_job_db_port(double prefix bug in the CouchDB bootstrap case)jqto runtime dependencies (needed for restore)DB_NAME/DB_NAME_EXCLUDEfor CouchDB in the READMETest plan
DB_NAME=testdbDB_NAME=ALL(system DBs excluded)DB_NAME=ALL+DB_NAME_EXCLUDE=somedbCloses #448