Skip to content

test: close the test and wording points that the reviews left open - #904

Merged
ThomasK33 merged 5 commits into
mainfrom
p3-docs-tests
Nov 17, 2026
Merged

ThomasK33 merged 5 commits into
mainfrom
p3-docs-tests

Conversation

@ThomasK33

@ThomasK33 ThomasK33 commented Nov 17, 2026 •

Copy link
Copy Markdown
Member

Closes the P3 points that the reviews of #897, #898, #900, #901 and #902 left open. No change in behavior: tests, docs and code comments only.

From Point Change
#897 No test read an empty-but-present OPENSTORAGE_SYNC_ALLOW_PRIVATE through serve's config: resolveSyncEgress gets a value-only lookup, so an empty value and an absent one are the same input there TestResolveServeConfigEmptySyncAllowPrivate (cmd/openstorage/config_test.go) runs resolveServeConfig with the variable empty (with and without the deprecated variable) and with false against a deprecated true. A mutant that reads an empty value as false at the call site fails it
#897 The docs called OPENSTORAGE_SYNC_PUSH_ALLOW_PRIVATE a "spelling" in some places and "a separate variable, not a legacy alias" in another docs/reference/configuration.md (table row, section intro), ADR-44, ADR-45, cmd/openstorage/README.md, and the comments in cmd/openstorage/config.go, wire.go, internal/sync/egress.go, worker.go and the boot warning test call it the deprecated push-only variable
#898 The missing-repository loop of TestReadValueListsPastGitArgBudget named no route in its failure, and its comment held only for the v1 routes Each case runs as a subtest named after its route (missing repository/diff, …); the comment says that the path-style alias looks the repository up first
#900 TestSyncValuesTheRunRefusesAtRegistration used one repository name for every POST case, so a case that stored a repository turned the next ones into repeat creates; a failure in its tracked-create part skipped the credential cases Each POST case mints a token for a repository name of its own; the tracked and repeat creates run as a subtest with t.Errorf
#900 The stored-text row of docs/explanation/wire-compatibility.md said "a base_repo.auth.token" for the line-break rule, which only a sync create applies "the base_repo.auth.token of a sync create"
#901 docs/how-to/deploy-production.md said the strict webhook policy fails closed unless an address is public and outside the IANA registries, but a public metadata endpoint (168.63.129.16) is denied too; the webhook false bullet of configuration.md filed the metadata endpoints under the IANA ranges; ADR-45 said the private classes hold every metadata endpoint; the guard.go note on 168.63.129.16 read as past tense Each sentence names the public endpoint
#902 The unit test of wcCheckRefPolicy wrapped the recorder with no envelope statuses, so it would keep passing if a commit write route started to answer 403 in its result envelope; internal/errtext/README.md did not list the unknown-op callers The test runs once per route in commitWriteEnvelopeStatuses; the README lists wrefPolicyCheck, wcCheckRefPolicy, wmCheckRefPolicy and the cut parse error of CheckRegistration

Two points of these reviews were fixed in their own PRs instead: the denied-address list of the sync true value (#901) and the "any length" comment on zones (#903).

Evidence

  • go test ./cmd/openstorage/ passes (60 s). go test -race -tags=openstorage_viewcheck of ./internal/events/... (22 s) and ./internal/sync/... (75 s) passes at 469bed4, and so does ./internal/api/... without pg (406 s). PostgreSQL and S3-WAL rely on CI.
  • vet, gofmt, the docs gate and both fingerprint gates pass (the worker.go change is a comment).

Review

  • 469bed4: the Codex review and the Codex security review found nothing. A fresh read-only review: PASS. The non-test Go changes are comments only; its four overlay mutants each fail the test that names them (an empty OPENSTORAGE_SYNC_ALLOW_PRIVATE read as false at resolveServeConfig; a 403 in restore-commit's envelope statuses; the blame list check after the repository lookup, which fails only missing repository/blame range; the reserved-prefix check dropped, which reports both creates and still runs the credential cases). The two merges of main kept everything.

Desk follow-up to the reviews of #897, #898, #900, #901 and #902.


Generated with xum • Model: anthropic:claude-opus-5-5 • Thinking: xhigh

…t open

- TestResolveServeConfigEmptySyncAllowPrivate: serve reads an
  OPENSTORAGE_SYNC_ALLOW_PRIVATE that is present but empty as unset, the
  one layer that can tell it from an absent one (review of #897).
- The docs and comments call OPENSTORAGE_SYNC_PUSH_ALLOW_PRIVATE the
  deprecated push-only variable, not a spelling: configuration.md also
  calls it a separate variable, not a legacy alias (review of #897).
- TestReadValueListsPastGitArgBudget runs each missing-repository case
  as a subtest named after its route, and its comment says that only the
  v1 routes check the list first (review of #898).
- TestSyncValuesTheRunRefusesAtRegistration gives each POST case a
  repository name of its own, and runs the tracked and repeat creates as
  a subtest, so one failure no longer hides the others (review of #900).
- The stored-text row of wire-compatibility.md scopes the line-break
  token rule to a sync create (review of #900).

Signed-off-by: Thomas Kosiewski <tk@coder.com>

---
_Generated with [`xum`](https://github.com/coder/xum) • Model: `anthropic:claude-opus-5-5` • Thinking: `xhigh`_

Change-Id: Ifbee7306f2c6802dfae82bfab4ea5e110cfa4291
Signed-off-by: Thomas Kosiewski <tk@coder.com>
Main gained #896 and #901. The only conflict is the table row of
OPENSTORAGE_SYNC_PUSH_ALLOW_PRIVATE: main names the metadata endpoints,
this branch calls the variable a variable; the row keeps both.

Signed-off-by: Thomas Kosiewski <tk@coder.com>

---
_Generated with [`xum`](https://github.com/coder/xum) • Model: `anthropic:claude-opus-5-5` • Thinking: `xhigh`_

Change-Id: I0b86a3e1b2048dc56b084e500e59e21d06c6ddc7
Signed-off-by: Thomas Kosiewski <tk@coder.com>
- deploy-production.md: the strict webhook policy also denies a cloud
  metadata endpoint that is a public address (Azure's 168.63.129.16).
- configuration.md: the webhook `false` bullet no longer files the
  metadata endpoints under the IANA special-purpose ranges.
- ADR-45: most metadata endpoints lie in the private classes, not all.
- guard.go: the note on 168.63.129.16 no longer reads as past tense.
- The rest of the code comments, ADR-44 and the boot warning test
  call OPENSTORAGE_SYNC_PUSH_ALLOW_PRIVATE a variable, not a spelling.

Signed-off-by: Thomas Kosiewski <tk@coder.com>

---
_Generated with [`xum`](https://github.com/coder/xum) • Model: `anthropic:claude-opus-5-5` • Thinking: `xhigh`_

Change-Id: I8e3b4beb6d4674c53f67f9500ce58b6a2c49d3f0
Signed-off-by: Thomas Kosiewski <tk@coder.com>
Change-Id: I0d4451f09f2f4493c9a9743abff43828811feed9
Signed-off-by: Thomas Kosiewski <tk@coder.com>
…op test

The review of #902 noted that the test of wcCheckRefPolicy wrapped the
recorder in an envelopeWriter with no statuses, so it would keep passing
if a commit write route later answered 403 in its result envelope. It
now runs once for each route in commitWriteEnvelopeStatuses. The errtext
README lists the unknown-op callers and the cut parse error of
CheckRegistration.

Signed-off-by: Thomas Kosiewski <tk@coder.com>

---
_Generated with [`xum`](https://github.com/coder/xum) • Model: `anthropic:claude-opus-5-5` • Thinking: `xhigh`_

Change-Id: I65a72822779feee84988fa6b8544a60a2aa509af
Signed-off-by: Thomas Kosiewski <tk@coder.com>
@ThomasK33

Copy link
Copy Markdown
Member Author

@codex review

@ThomasK33

Copy link
Copy Markdown
Member Author

@codex security review

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Nov 17, 2026 •

Copy link
Copy Markdown
Codex Review Summary: ✅ Completed

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-11-17T00:10:57.014224Z 469bed4 Manual request
🔒 Security Review ✅ Completed 2026-11-17T00:12:23.193098Z 469bed4 Manual request
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. You're on a roll.

Reviewed commit: 469bed430a

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@chatgpt-codex-connector

Copy link
Copy Markdown

🛡️ Codex Security Review

Security review completed. No security issues were found in this pull request.

Reviewed commit: 469bed430a

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

  • You comment "@codex security review"
  • A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

@ThomasK33
ThomasK33 added this pull request to the merge queue Nov 17, 2026
Merged via the queue into main with commit 3f23b1c Nov 17, 2026
17 checks passed
@ThomasK33
ThomasK33 deleted the p3-docs-tests branch November 17, 2026 01:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant