Fix: Sometimes `manage_queues` does not dump all available messages (#3663) by hannes-ucsc · Pull Request #8414 · DataBiosphere/azul

hannes-ucsc · 2026-10-17T22:51:23Z

Linked issues: #3663

Checklist

Author

¹ when the issue title describes a problem, the corresponding PR
title is Fix: followed by the issue title

Author (partiality)

B01 Added p tag to titles of partial commits
B02 This PR is labeled partial _{or completely resolves all linked issues}
B03 This PR partially resolves each of the linked issues _{or does not have the partial label}

Author (reindex)

C01 Added r tag to commit title _{or the changes introduced by this PR will not require reindexing of any deployment}
C02 This PR is labeled reindex:dev _{or the changes introduced by it will not require reindexing of dev}
C03 This PR is labeled reindex:anvildev _{or the changes introduced by it will not require reindexing of anvildev}
C04 This PR is labeled reindex:anvilprod _{or the changes introduced by it will not require reindexing of anvilprod}
C05 This PR is labeled reindex:prod _{or the changes introduced by it will not require reindexing of prod}
C06 This PR is labeled reindex:partial and its description documents the specific reindexing procedure for dev, anvildev, anvilprod and prod _{or requires a full reindex or carries none of the labels reindex:dev, reindex:anvildev, reindex:anvilprod and reindex:prod}

Author (mirror)

D01 This PR is labeled mirror:dev _{or the changes introduced by it will not require mirroring of dev}
D02 This PR is labeled mirror:anvildev _{or the changes introduced by it will not require mirroring of anvildev}
D03 This PR is labeled mirror:anvilprod _{or the changes introduced by it will not require mirroring of anvilprod}
D04 This PR is labeled mirror:prod _{or the changes introduced by it will not require mirroring of prod}
D05 This PR is labeled mirror:partial and its description documents the specific mirroring procedure for dev, anvildev, anvilprod and prod _{or requires a full mirroring or carries none of the labels mirror:dev, mirror:anvildev, mirror:anvilprod and mirror:prod}

Author (API changes)

E01 This PR and its linked issues are labeled API _{or this PR does not modify a REST API}
E02 Added a (A) tag to commit title for backwards (in)compatible changes _{or this PR does not modify a REST API}
E03 Updated REST API version number in app.py _{or this PR does not modify a REST API}

Author (upgrading deployments)

F01 Ran make docker_images.json and committed the resulting changes _{or this PR does not modify azul_docker_images, or any other variables referenced in the definition of that variable}
F02 Documented upgrading of deployments in UPGRADING.rst _{or this PR does not require upgrading deployments}
F03 Added u tag to commit title _{or this PR does not require upgrading deployments}
F04 This PR is labeled upgrade _{or does not require upgrading deployments}
F05 This PR is labeled deploy:shared _{or does not modify docker_images.json, and does not require deploying the shared component for any other reason}
F06 This PR is labeled deploy:gitlab _{or does not require deploying the gitlab component}
F07 This PR is labeled deploy:runner _{or does not require deploying the runner image}

Author (hotfixes)

G01 Added F tag to main commit title _{or this PR does not include permanent fix for a temporary hotfix}
G02 Reverted the temporary hotfixes for any linked issues _{or the none of the stable branches (anvilprod and prod) have temporary hotfixes for any of the issues linked to this PR}

Author (before every review)

H01 Rebased PR branch on develop, squashed fixups from prior reviews
H02 Ran make requirements_update _{or this PR does not modify pyproject.toml}
H03 Added R tag to commit title _{or this PR does not modify uv.lock}
H04 This PR is labeled reqs _{or does not modify uv.lock}
H05 make integration_test passes in personal deployment _{or this PR does not modify functionality that could affect the IT outcome}
H06 PR is awaiting requested review from a peer
H07 Status of PR is Review requested
H08 PR is assigned to only the peer and the author

Peer reviewer (after approval)

Note that after requesting changes, the PR must be assigned to only the author.

J01 Actually approved the PR
J02 PR is not a draft
J03 PR is awaiting requested review from system administrator
J04 Status of PR is Review requested
J05 PR is assigned to only the system administrator and the author

System administrator (after approval)

K01 Actually approved the PR
K02 Labeled linked issues as demo or no demo
K03 Commented on linked issues about demo expectations _{or all linked issues are labeled no demo}
K04 Decided if PR can be labeled no sandbox
K05 A comment to this PR details the completed security design review
K06 PR title is appropriate as title of merge commit
K07 N reviews label is accurate
K08 Status of PR is Approved
K09 PR is assigned to only the operator and the author

Operator

L01 Checked reindex:… labels and r commit title tag
L02 Checked mirror:… labels
L03 Checked that demo expectations are clear _{or all linked issues are labeled no demo}
L04 Squashed PR branch and rebased onto develop
L05 Sanity-checked history
L06 Pushed PR branch to GitHub

Operator (deploy `.shared` and `.gitlab` components)

M01 Ran _select dev.shared && CI_COMMIT_REF_NAME=develop make -C terraform/shared apply_keep_unused _{or this PR is not labeled deploy:shared}
M02 Ran _select dev.gitlab && CI_COMMIT_REF_NAME=develop make -C terraform/gitlab apply(an error from _login_docker_gitlab is benign if the instance was stopped for backup) _{or this PR is not labeled deploy:gitlab}
M03 Ran _select anvildev.shared && CI_COMMIT_REF_NAME=develop make -C terraform/shared apply_keep_unused _{or this PR is not labeled deploy:shared}
M04 Ran _select anvildev.gitlab && CI_COMMIT_REF_NAME=develop make -C terraform/gitlab apply(an error from _login_docker_gitlab is benign if the instance was stopped for backup) _{or this PR is not labeled deploy:gitlab}
M05 Checked the items in the next section _{or this PR is labeled deploy:gitlab}
M06 PR is assigned to only the system administrator and the author _{or this PR is not labeled deploy:gitlab}

System administrator (post-deploy of `.gitlab` component)

N01 Background migrations for dev.gitlab are complete _{or this PR is not labeled deploy:gitlab}
N02 Background migrations for anvildev.gitlab are complete _{or this PR is not labeled deploy:gitlab}
N03 PR is assigned to only the operator and the author

Operator (deploy runner image)

P01 Ran _select dev.gitlab && make -C terraform/gitlab/runner _{or this PR is not labeled deploy:runner}
P02 Ran _select anvildev.gitlab && make -C terraform/gitlab/runner _{or this PR is not labeled deploy:runner}

Operator (sandbox build)

Operator (merge the branch)

R01 All status checks passed and the PR is mergeable
R02 The title of the merge commit starts with the title of this PR
R03 Added PR # reference to merge commit title
R04 Collected commit title tags in merge commit title _{but only included p if the PR is also labeled partial}
R05 Pushed merge commit to GitHub
R06 Status of PR is Merged lower
R07 Status of blocked issues is Triage _{or no issues are blocked on the linked issues}

Operator (main build)

Operator (reindex)

Operator (mirroring)

U01 Started mirroring in dev _{or this PR is not labelled mirror:dev}
U02 Started mirroring in anvildev _{or this PR is not labelled mirror:anvildev}
U03 Checked for, triaged and possibly requeued messages in mirror fail queue in dev _{or this PR is not labelled mirror:dev}
U04 Checked for, triaged and possibly requeued messages in mirror fail queue in anvildev _{or this PR is not labelled mirror:anvildev}
U05 Emptied mirror fail queue in dev _{or this PR is not labelled mirror:dev}
U06 Emptied mirror fail queue in anvildev _{or this PR is not labelled mirror:anvildev}

Operator

V01 Propagated the upgrade, API, deploy:shared, deploy:gitlab, deploy:runner, reindex:partial, reindex:anvilprod, reindex:prod, mirror:partial, mirror:anvilprod and mirror:prod labels to any open promotion PRs _{or this PR carries none of these labels, or is not included in an open promotion PR}
V02 Propagated any specific instructions related to those labels, from the description of this PR to that of any open promotion PRs _{or this PR carries none of those labels, or is not included in an open promotion PR}
V03 PR is assigned to no one

Shorthand for review comments

L line is too long
W line wrapping is wrong
Q bad quotes
F other formatting problem

codecov · 2026-10-17T23:00:53Z

Codecov Report

❌ Patch coverage is 90.32258% with 39 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.42%. Comparing base (8842489) to head (9c31eed).

Files with missing lines	Patch %	Lines
src/azul/queues.py	79.21%	37 Missing ⚠️
src/azul/deployment.py	50.00%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #8414      +/-   ##
===========================================
+ Coverage    85.00%   85.42%   +0.41%     
===========================================
  Files          170      171       +1     
  Lines        26085    26423     +338     
===========================================
+ Hits         22174    22571     +397     
+ Misses        3911     3852      -59

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

coveralls · 2026-10-17T23:01:26Z

coverage: 85.489% (+0.4%) from 85.078% — issues/hannes-ucsc/3663-manage-queues into develop

hannes-ucsc · 2026-10-22T03:44:00Z

The entity types in an AnVIL bundle are the names of the BigQuery tables the entities are drawn from, and therefore carry an `anvil_` prefix. The set of types to redact lists them without that prefix, so the membership test never holds, and `--redaction-key` quietly writes the bundle out unredacted. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A bundle holds its orphans separately from its entities, and only the latter were redacted, leaving an orphaned biosample, diagnosis or donor in the clear. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The SQS redrive facility moves messages server-side, so unlike the dump and feed commands it is not subject to the limitations of receiving from a FIFO queue, where the messages of a group behind an in-flight head remain unreachable. The source queue must be configured as the dead-letter queue of another queue, and both queues must be of the same type. The command waits for the move to complete. Since SQS only reports the handle of a task while that task is running, a finished task is identified by when it started. Also make the optional arguments of the client method keyword-only, matching how every call site already passes them. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The check reads the queue property of the dump envelope, which a later commit replaces. Validation returns at the end of this branch, in terms of the type of message rather than the identity of the queue, so that feeding the contents of a fail queue into the corresponding work queue no longer requires an override. The --force flag goes with the check, and the procedure in OPERATOR.rst no longer refers to force-feeding. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

_dump_messages condensed the messages it was given, but the cleanup of an interrupted feed passes it the already condensed messages it read from the dump file. The resulting AttributeError was raised while handling the exception that interrupted the feed, hiding the actual cause of the interruption from the operator. The messages are now condensed by _dump, the only caller that holds uncondensed ones. Present since 2cf70b9. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The cleanup of an interrupted feed rewrote the dump file from the complete list of messages instead of the ones that had not been sent yet, so feeding the file again sent those messages a second time. The list of remaining messages was computed but only used to decide whether any message had been sent at all. Present since 2cf70b9. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Neither command was covered by a test. One test round-trips the messages of a queue through a dump file, the other interrupts a feed and covers the two defects fixed in dcf0d854f and eea840207, by asserting that the exception which interrupted the feed is the one that propagates, and that the file is left holding exactly the messages that were not sent. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Appending to a dump file is a prerequisite for deleting each batch of messages as it is written, which is in turn what allows a FIFO queue to be dumped completely. This commit only changes the format, leaving the dump to collect every message in memory before writing it, and the feed to read the whole file into memory. The first line of a dump is the envelope, each subsequent line is one message. Dumps written by earlier versions are no longer accepted, and dump_all now names its files after the queue with a .jsonl suffix. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A message held in flight blocks the rest of its group, so a FIFO queue cannot be dumped completely unless each batch is deleted from the queue before the next one is received. Each batch is now written to the dump file and flushed to disk before it is deleted, so that an interrupted dump leaves behind every message it removed from the queue. This also keeps the number of in-flight messages at one batch, well below the limit SQS imposes on a queue, and it bounds the time a message is invisible by the duration of one batch instead of the whole dump. Because the messages of a dump that does not delete them must be held in flight until the dump ends, or receiving them again would make the dump run forever, such a dump still captures only the messages at the head of each group. It now says so before it starts. The dump file is no longer written through write_file_atomically, whose removal of the temporary file on failure would discard exactly those messages that were already deleted from the queue. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A short poll samples a subset of the SQS hosts and can therefore return nothing while messages remain, ending a dump prematurely. This affects FIFO and standard queues alike. A long poll queries all hosts, and the documented maximum wait is used because AWS notes that lower values make empty responses more likely. The wait is an attribute so that tests can disable it. Moto honours the parameter by blocking, and every drain ends on an empty receive, which would otherwise add twenty seconds to each one. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The wait after purging a queue added the number of available messages twice and never counted the ones in flight, so it could return while messages were still being processed, just before the readers of the queue are enabled again. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

An empty receive ended a dump even when the queue still reported messages. The dump now consults those counts, discounting the messages it holds in flight itself, and keeps polling while they disagree with the empty receive. The counts are approximate, and the producers of a queue keep adding to it unless they are disabled, so polling is bounded: after three consecutive empty receives the dump gives up and says what the queue still reports. The names of the three counting attributes are now defined once, and the wait after purging a queue uses them, too. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The --no-json-body flag stored the body of a message verbatim instead of embedding the JSON it contains. Nothing but the command line ever set it, and every message azul queues carries a JSON body, so a body that the flag was needed for could only come from outside azul. The inverse operation loses the branch that re-serialized such a body, which means a dump written with the flag can no longer be fed back. Dumps written before this branch could not be, either, since the format has changed. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The class is not slotted because its cached property caches by writing to the instance dictionary, which a slotted class lacks. The sole field keeps its leading underscore, so that attrs names the constructor argument after it without one, as the callers already do. Only the constant naming the attributes that count the messages in a queue is annotated, so it is the only one attrs would have mistaken for a field, and it is declared as a class variable to prevent that. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The two descriptions differed only in whether they spoke of one queue and file or of several, which the wording now avoids. One of them also said "the its file". Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The two subcommands defined the option identically, which left them free to drift apart. They now inherit it from a parser that carries nothing else. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The lambdas reading the work queues produce the messages in every queue, so a dump taken while they are running is likely to miss messages. The dump now consults the work queues, and if they hold messages, insists on a decision: --safe disables those lambdas for its duration, --unsafe accepts a dump that is probably incomplete. Combining --unsafe with --delete is refused, because it would discard the messages the dump missed. A dump taken while the work queues are empty needs no decision, and only warns that nothing kept messages from arriving while it ran. The two flags write to one tri-state attribute, so they cannot both be in effect. The mutually exclusive group they share only improves the error message. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

An interrupted feed rewrote the dump with the messages it had not submitted, which both destroyed the dump it was fed from and left nothing to say how far it had come. It now records that in a file beside the dump, rewritten after every batch, and skips as many messages when it is run again. The file names the queue it fed, so that a dump interrupted on its way into one queue is not silently truncated on its way into another. The --delete flag now removes the dump once every message in it was submitted, instead of removing them from it one batch at a time. A crash between submitting a batch and recording it resubmits that batch when the feed is resumed. On a FIFO queue, SQS discards the resulting duplicates, because the deduplication IDs are carried over from the dump. On a standard queue it does not. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A dump now records the type of message it holds, and a feed refuses a dump whose type is not the one the target queue is expected to hold. There are three mutually incompatible types, each named after the class whose instances are the bodies of those messages. Those classes are defined in modules that import this one, hence the deferred import. Unlike the check this replaces, which compared the queue a dump was taken from against the one it is fed into, this one permits feeding the contents of a fail queue into the work queue it belongs to, which is what the operator manual prescribes. It has no override, because no queue can do anything with a message of another type. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A dump records the queue it was taken from, and a feed refuses a dump whose deployment is not the current one, unless --foreign says to proceed. Type compatibility does not catch this, because the messages of one deployment are perfectly well-formed messages of another. Moving messages between deployments is a legitimate thing to do, for instance to reproduce a failure in a personal deployment, which is why this check has an override while the one on the type of message does not. The queue is recorded by name rather than by URL, because the name is what carries the deployment, and the URL adds an account and a region to a file that operators read. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

A dump used to replace the contents of the file it was given, silently discarding an earlier dump. It now refuses that file unless --overwrite says to replace it, or --append says to add to it. Appending compares the header of the file against the one the dump would write, so that a dump of one queue is not appended to a dump of another. The two flags write to one tri-state attribute, as the flags deciding what a dump does about the queue worker lambdas do. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Every condition the queue management refuses to work under is a demand it makes of its input or of the state of the deployment, not a defect in itself, which is what R is for. Two of them were stated as branches of a chain that had no condition to assert, and are now stated as the conditions under which a dump of a busy deployment, and a dump to a file that exists, may proceed. The chain deciding whether to append collapses as a result, the attribute being a plain boolean once it is known not to be None. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

The procedure spoke of notifications and had the operator repeat it for tallies afterwards, as an aside. It now covers both at once, in numbered steps, and prescribes feeding the tallies first, they being downstream from the notifications. Between dumping and feeding, the operator is now asked to triage the failures and to agree on a remedy, rather than to feed the messages back as they are. Messages are now fed into the work queue, with the retry queue reserved for the case where most of them are known to need the memory and the timeout of the retry lambda, instead of being fed into the retry queue to save time. The constraints the script enforces on either step are described as well, along with the flags that relax those that may be relaxed. The instructions for tracking the remaining failures in tickets are removed. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Indexing a bare dict yields Any, which spreads to the body and the URL of an existing PR. Typing the response as JSON makes that indexing yield a JSON value, which json_str narrows to the string both of them are. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

hannes-ucsc self-assigned this Oct 17, 2026

hannes-ucsc linked an issue Oct 17, 2026 that may be closed by this pull request

Sometimes manage_queues does not dump all available messages #3663

Open

hannes-ucsc force-pushed the issues/hannes-ucsc/3663-manage-queues branch from 396f7c4 to 208aa31 Compare October 22, 2026 00:40

hannes-ucsc and others added 24 commits October 21, 2026 21:56

Redact the orphans of an AnVIL bundle, too

45a55e1

A bundle holds its orphans separately from its entities, and only the latter were redacted, leaving an orphaned biosample, diagnosis or donor in the clear. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Describe --delete the same way for dump and dump_all

2e3360d

The two descriptions differed only in whether they spoke of one queue and file or of several, which the wording now avoids. One of them also said "the its file". Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Share the --delete option between dump and dump_all

1256b93

The two subcommands defined the option identically, which left them free to drift apart. They now inherit it from a parser that carries nothing else. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Call the first line of a dump its header

97bad7f

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

hannes-ucsc and others added 25 commits October 21, 2026 21:56

Fix long lines

e217ba8

Rename variable

210f96d

Move method call

375ec1d

Replace attrs with a constructor

7221319

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Reorder methods

8545fa4

Rename members

fae59fd

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Convert global to local variables

1701416

Wrap long lines

71a71e6

Rename methods

b9233b6

Eliminate inconsistent use of "item" vs "task"

f1ca872

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Rename method

ff43d22

Reorder members

20c412e

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Inline methods

3091495

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Extract method

22774a4

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Inline methods

9c61004

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Inline methods

0072d26

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Fix odd line wrapping

db32050

Inline method

724394b

Inline and extract methods

3f41be6

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Inline method and rename local variables

03cb875

Rename methods

131a7cb

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Inline method

ab6abde

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Rename property

c7a433a

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Reorder members

9c31eed

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

hannes-ucsc force-pushed the issues/hannes-ucsc/3663-manage-queues branch from 208aa31 to 9c31eed Compare October 22, 2026 06:05

hannes-ucsc added the sandbox [process] Resolution is being verified in sandbox deployment label Oct 22, 2026

hannes-ucsc merged commit 2a07043 into develop Oct 22, 2026
10 checks passed

hannes-ucsc deleted the issues/hannes-ucsc/3663-manage-queues branch October 22, 2026 20:20

hannes-ucsc removed their assignment Oct 22, 2026

Navigation Menu

Repository navigation

Fix: Sometimes `manage_queues` does not dump all available messages (#3663) - #8414

Fix: Sometimes `manage_queues` does not dump all available messages (#3663)#8414
hannes-ucsc merged 95 commits into
developfrom
issues/hannes-ucsc/3663-manage-queues

hannes-ucsc commented Oct 17, 2026 •

edited

Loading

Uh oh!

codecov Bot commented Oct 17, 2026 •

edited

Loading

Uh oh!

coveralls commented Oct 17, 2026 •

edited

Loading

Uh oh!

hannes-ucsc commented Oct 22, 2026

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

hannes-ucsc commented Oct 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Checklist

Author

Author (partiality)

Author (reindex)

Author (mirror)

Author (API changes)

Author (upgrading deployments)

Author (hotfixes)

Author (before every review)

Peer reviewer (after approval)

System administrator (after approval)

Operator

Operator (deploy .shared and .gitlab components)

System administrator (post-deploy of .gitlab component)

Operator (deploy runner image)

Operator (sandbox build)

Operator (merge the branch)

Operator (main build)

Operator (reindex)

Operator (mirroring)

Operator

Shorthand for review comments

Uh oh!

codecov Bot commented Oct 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

coveralls commented Oct 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

hannes-ucsc commented Oct 22, 2026

Security design review

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

hannes-ucsc commented Oct 17, 2026 •

edited

Loading

Operator (deploy `.shared` and `.gitlab` components)

System administrator (post-deploy of `.gitlab` component)

codecov Bot commented Oct 17, 2026 •

edited

Loading

coveralls commented Oct 17, 2026 •

edited

Loading