Fix dry-run to look up mapping files by secret ID, not name#28
Merged
Conversation
Podman's shell driver stores mapping files at state_dir/{SECRET_ID}
where SECRET_ID is the hex identifier, not the human-readable name.
The dry-run classification was checking state_dir/{name} which never
matched, causing every secret to be reported as orphaned even when
mapping files existed.
Discovered when running psi setup --dry-run on a production host:
519 mapping files existed on disk but all 514 shell-driver secrets
were classified as orphaned because the filename is a hex ID like
00e33d7a267c46aea501654b0, not atuin--ATUIN_DB_URI.
Read the ID field from the Podman API response and use it for the
state_dir lookup.
jdoss
added a commit
that referenced
this pull request
Apr 16, 2026
Setup populated the encrypted cache with keys like "atuin--POSTGRES_PASSWORD" but the serve lookup path receives the Podman hex secret ID in $SECRET_ID (e.g. "40ba10c50224daba166308fbc"). The cache.get() call in serve never matched, so the cache never served a single hit — every lookup always went to the live provider. This is the same class of bug as the dry-run fix in PR #28: Podman identifies secrets by hex ID internally, not by human-readable name. The cache exists specifically to serve secrets when the upstream provider is down. In its first real-world test (Infisical down during a container restart cascade), it failed silently — every Infisical- backed lookup returned a provider error instead of the cached value. Fix: _register_secrets now returns a {key: hex_id} map from the Podman API's create response. The caller uses hex IDs as cache keys so serve's cache.get(secret_id) finds the right entry.
3 tasks
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
1 participant
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
Podman's shell driver stores mapping files at
state_dir/{SECRET_ID}(hex identifier), notstate_dir/{name}. The dry-run classification was checkingstate_dir/{name}, which never matched — every secret was reported as orphaned even when mapping files existed on disk.Discovered when running
psi setup --dry-runon a test server: 519 mapping files existed but all 514 shell-driver secrets were classified as orphaned.Fix: read the
IDfield from the Podman API response and use it for thestate_dirlookup.Test plan
pytest tests/test_dry_run_setup.py— all tests updated to usesecret_idfor mapping file names.ruff check/ruff format --check/ty check— clean.