Datasets:

giseldo

/

neodataset

like 1

2020-02-17 18:54:11.122

Execution order for Label Inference and Triage Report

## Problem The order of Category Label Inference and Triage report is causing some confusion and I'm not able to easily track down what ran when with the large amount of pipelines. . From @jyavorska on [Slack](https://gitlab.slack.com/archives/C3JJET4Q6/p1581932656014900): >>> Hello, possibly it's still expected but I noticed the bots ran out of order again - I received triage reports of the need to apply category labels that another bot came along and automatically applied category labels to. (edited) >>> - [ ] What is the order for the pipelines on 2020-02-16 on `gitlab-org` - Label Inference daily: https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/439127366 - Team Triage weekly: https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/439125366 - [ ] Is there a way to simplify this or ensure that triage packages are created after the label inference for that week.

2020-02-14 07:36:37.934

Add group name prefix to job names

Add prefix `gitlab-org` to the following jobs: - `label-merge-request-author-group` - `stage-and-group-labels-hygiene` Follow up from https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/390#note_287596184)

2020-02-11 03:07:49.198

Draft a plan for an escalation chain for ~availability S1, S2, and S3 have exceeded the time to resolve

## Proposal * [ ] Draft a plan for an escalation chain when ~S1, ~S2, and ~S3 have exceeded the time to resolve.

2020-02-11 03:00:47.254

Apply ~bug to ~availability issues

## Proposal - [ ] Add a hygiene rule for the gitlab-org group level which will apply ~bug for all issues labeled ~availability without ~bug. * See @meks reasoning here: > As I stated before in https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_requests/39270#note_280407345, ~availability issues are a category of bugs, we don't have a good hygiene on label yet (which we will fix with automation). This iteration aims to have better clarity and specifics on how to deal with bugs/defects that immediately impacts GitLab.com ~availability. The overall ~bug pie is huge, this sets out to make a subset of that better with clearer and specific actionable steps.

2020-02-11 02:53:36.586

Add new section to Triage Report for ~"availability" issues

## Proposal - [ ] Add a ~availability Issues section to the triage report above the ~bug section. - [ ] List issue by severity/priority for all group + ~availability issues

2020-02-08 00:33:29.024

Label automation enforcement for ~availability issues in product facing projects

Per the new ~availability Priorization and Severity scheme https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_requests/39270 we should do a few triage-ops automation to help enforce and raise awareness. ### Triage report - [x] Add a new section above ~bug to list issues that are labelled with ~availability - https://gitlab.com/gitlab-org/quality/triage-ops/issues/406 ### Triage Automation - [x] Label issues that have ~availability with ~bug. ~availability is another category of bugs. - https://gitlab.com/gitlab-org/quality/triage-ops/issues/407 - See my reasoning here: - > As I stated before in https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_requests/39270#note_280407345, ~availability issues are a category of bugs, we don't have a good hygiene on label yet (which we will fix with automation). This iteration aims to have better clarity and specifics on how to deal with bugs/defects that immediately impacts GitLab.com ~availability. The overall ~bug pie is huge, this sets out to make a subset of that better with clearer and specific actionable steps. - [x] Enforce [prioritization band](https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_requests/39270#prioritization-band) with Triage Serverless. - https://gitlab.com/gitlab-org/quality/triage-serverless/issues/22 - When an issue with ~availability has a Severity label, automatically add the minimally allowed Priority label. - ~availability ~S1 => ~P1 - ~availability ~S2 => ~P1 - ~availability ~S3 => ~P2 - ~availability ~S4 => ~P3 - Enforce prioritization, we do not allow the following labels - ~availability ~S1 cannot have ~P2 ~P3 ~P4 it can only have ~P1 - ~availability ~S2 cannot have ~P2 ~P3 ~P4 it can only have ~P1 - ~availability ~S3 cannot have ~P3 ~P4 it can only have ~P2 or ~P1 (extra credit) - ~availability ~S4 cannot have ~P4 it can only have ~P3 or ~P2 / ~P1 (extra credit) - [x] Draft a plan for an escalation chain when ~S1, ~S2, and ~S3 have exceeded the time to resolve. - https://gitlab.com/gitlab-org/quality/triage-ops/-/issues/426 - [x] Add dashboards to visualize - https://gitlab.com/gitlab-org/gitlab-insights/issues/161 ---- | Severity | Availability Impact | Reproducibility | Time to resolve | Deployment target | Minimum priority | | -------- | ------- | -------------- | --------------- | ----------------- | ----------------- | | ~S1 | Roadblock on GitLab.com and blocking customer's business goals and day-to-day workflow | Consistently reproducible | Within 48 hrs | Hotfix to GitLab.com | ~P1 | | ~S2 | Significant impact on GitLab.com and customer's day-to-day workflow. Customers have an acceptable workaround in place. | Consistently reproducible | Within 5 business days | Next deployment window after resolution | ~P1 | | ~S3 | Broad impact on GitLab.com and minor inconvenience to customer's day-to-day workflow. No workaround needed. | Inconsistently reproducible | Within 30 days | Next release after resolution | ~P2 | | ~S4 | Minimal impact on GitLab.com, no known customers affected | Inconsistently reproducible | Within 60 days | Next release after resolution | ~P3 | Merge request https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/39270/diffs ## Prioritization band To call out specifics on what priorities can be set on an ~availability issue, please refer to the table below. As we implement these labels, we can have triage automation enforce these priorities based on the labels set forth. | Issue | Allowed Priorities | | ------ | ------ | | ~availability ~S1 | ~P1 only | | ~availability ~S2 | ~P1 only | | ~availability ~S3 | ~P2 as baseline, ~P1 allowed | | ~availability ~S4 | ~P3 as baseline, ~P2 ~P1 allowed |

2020-02-03 04:32:52.639

Use Issue and Merge Request resource class to determine how to infer the labels

Previously the method `#new_stage_and_group_labels_from_intelligent_inference` uses `infer_from_category: false` for merge request label inferrence. With https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/390#note_281173182, the inference between an Issue and Merge Request are becoming orthogonal. I'm suggesting to change from using `infer_from_category` flag to determine whether to infer for an MR or an issue to using the resource class itself.

2020-02-03 03:31:40.683

Add comment mentioning that the author's correct group could not be determined

Following from https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/390#note_278719721, some MRs may not have group label applied because the author's correct group could not be determined. This may happen in the following scenarios: - MR author is a gitlab team member: - The MR author's group is not listed in `www-gitlab-com` `data/team.yml` under `departments` - The MR author may belong to more than one teams (e.g member of one team and intern in another) - MR author is a community member /cc @gl-quality/eng-prod

2020-01-31 10:37:54.874

Cut down on assignees on group triage report issues

### Summary Sometimes group triage report issues are created but do not need to cover a certain topic. For example, all frontend bugs are scheduled so there is no section for frontend bugs. In this case we still assign the Frontend Engineering Manager to the triage report. Let's make it so that people that do not need to be involved in a triage report are not assigned ### Aims - Cut down on the number of triage reports that people are unnecessarily assigned to - Hopefully this will make triage reports less annoying and improve engagement ### Proposal Instead of using a collection of assignees, use a Hash by department type and add a quick action to assign the relevant users as part of their summary action

2020-01-29 11:21:58.542

Delete "[DEPRECATED]" label

### Description Hi there! This is Arturo from the Ecosystem team (cc. @nhxnguyen). We like to remove one label: team-tasks~10375644 because we are now using team-tasks~10690808. I've searched for the issues and merge requests with that label and we can safely remove the label, see: - https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=Ecosystem%20%5BDEPRECATED%5D - https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=Ecosystem%20%5BDEPRECATED%5D Can you help us with that, please? ### Plan We have deleted ~"Ecosystem [DEPRECATED]", we are currently evaluating if anything broke, if nothing adverse was an effect, we can start deleting all of them. - [-] Create a record of issues which do not have a go forward devops:: label based on legacy team label: https://gitlab.com/gitlab-org/quality/triage-ops/issues/399#note_293402819 - [x] Remove team labels from label inference for Triage Ops: https://gitlab.com/gitlab-org/quality/triage-ops/-/blob/47ee6d33bde3eb142dbd11bf7918037660ee8a0b/lib/devops_labels.rb#L20-41 => https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/432 - [x] Remove Deprecated team dashboards from the Quality Dashboard: https://quality-dashboard.gitlap.com/groups/gitlab-org/teams => https://gitlab.com/gitlab-org/gitlab-insights/-/merge_requests/191 - [x] Delete the following labels - [x] ~"CI/CD [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=CI%2FCD%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=CI%2FCD%20%5BDEPRECATED%5D)) - [x] ~"Distribution [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Distribution%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Distribution%20%5BDEPRECATED%5D)) - [x] ~"Gitaly [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Gitaly%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Gitaly%20%5BDEPRECATED%5D)) - [x] ~"Memory [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Memory%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Memory%20%5BDEPRECATED%5D)) - [x] ~"Gitter [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Gitter%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Gitter%20%5BDEPRECATED%5D)) - [x] ~"Growth [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Growth%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Growth%20%5BDEPRECATED%5D)) - [x] ~"Monitor [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Monitor%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Monitor%20%5BDEPRECATED%5D)) - [x] ~"Package [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Package%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Package%20%5BDEPRECATED%5D)) - [x] ~"Release [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Release%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Release%20%5BDEPRECATED%5D)) - [x] ~"Serverless [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Serverless%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Serverless%20%5BDEPRECATED%5D)) - [x] ~"Defend [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Defend%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Defend%20%5BDEPRECATED%5D)) - [x] ~"Secure [DEPRECATED]" ([0 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Secure%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Secure%20%5BDEPRECATED%5D)) - [x] ~"Platform [DEPRECATED]" ([738 open issues](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=Platform%20%5BDEPRECATED%5D), [0 open MRs](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=opened&label_name[]=Platform%20%5BDEPRECATED%5D)) => can be safely removed since it was split into Manage, Plan, Create. - [x] ~"Verify [DEPRECATED]" - [x] ~"Configure [DEPRECATED]" - [ ] Get https://gitlab.com/gitlab-org/quality/triage-ops/issues/242 done and delete: - [ ] ~"Create [DEPRECATED]" - [ ] ~"Plan [DEPRECATED]" - [ ] ~"Manage [DEPRECATED]"

2020-01-27 18:59:43.338

Review Communiuty contributions MR report for 01-27-2020

The weekly report only ran for Runner this week, and I didn't see a report for `gitlab-org` or `gitaly` Relevant info: https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/415424122

2020-01-21 12:09:10.570

Group or stage report for flaky-master specs

There's a concern that quarantined flaky-master specs could stick around. Should we create a report around flaky-master specs that have passed a certain threshold? ## Proposal - Add a section to the team-triage report that lists ~P1 ~"failure::flaky-test" non-~QA issues - ~QA issues are triaged by the Quality Triage Report

2019-12-24 15:16:17.080

`dry-run:schedule` for now will only run the first schedule from the variants

I realized this when I was thinking how to dry-run some of the schedules. The problematic code is in `Schedule::Manager#find_schedule!`: ``` ruby def find_schedule!(project_path) schedules_specification.find do |spec| spec.project_path == project_path end || raise( InvalidDryRunScheduleName, "Cannot find schedule named #{project_path.inspect} in #{raw_schedules_path}") end ``` It's using `find` to find the first matching specification, however, there can have multiple ones with the same `project_path`. `find` is a linear search, so we always return the first one. For example, in the current schedule `gitlab-org/gitaly`, there are two schedules underneath: ``` yaml gitlab-org/gitaly: base: variables: TRIAGE_SOURCE_TYPE: projects TRIAGE_SOURCE_PATH: 2009901 variants: - id: 29054 # https://gitlab.com/gitlab-org/quality/triage-ops/pipeline_schedules/29054/edit active: true description: '[DAILY] gitlab-org/gitaly' variables: # hygiene TRIAGE_LABEL_ACCEPTING_MERGE_REQUESTS: 1 TRIAGE_LABEL_MISSED_SLO: 1 TRIAGE_LABEL_REMINDERS: 1 TRIAGE_MOVE_MILESTONE_FORWARD: 1 - id: 29055 # https://gitlab.com/gitlab-org/quality/triage-ops/pipeline_schedules/29055/edit active: true description: '[WEEKLY] gitlab-org/gitaly' cron: '0 0 * * 1' variables: # package TRIAGE_COMMUNITY_MERGE_REQUESTS: 1 TRIAGE_MISSING_CATEGORIES: 1 TRIAGE_TEAM_TRIAGE_PACKAGE: 1 ``` For now it'll only dry-run the first one (daily), ignoring the second one (weekly) This might not be important to fix before we actually need it, but we need to know this limitation. /cc @gl-quality/eng-prod

2019-12-20 03:48:39.466

Apply default labels for gitlab-org/charts/gitlab-runner

## Proposal Apply runner related labels to Issues in https://gitlab.com/gitlab-org/charts/gitlab-runner similar to https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/projects/gitlab-runner/default-labels.yml - ~"devops::verify" - ~"group::runner" - ~"Category:Runner"

2019-12-16 18:26:32.694

Update auto-labeling from the bot to account for new categories

* Groups are now called Subgroups https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/36874 * Category updates for Manage:Compliance - Audit Events, Audit Management, and Compliance Controls https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/34261 * Static site editor to Create https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/32938 * Update Secure Stage with new Group and Categories https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/35204 * Update Defend Stage with new group "Anomaly Detection" and moving categories to correct group https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/35196 cc @gl-quality/eng-prod

2019-11-26 15:41:38.689

Feedback on Unlabelled Issues Triage Package

## Summary During the [EP Team meeting](https://docs.google.com/document/d/1QxRQRLP-GpiCis4oFvm0NLm1lJm6WcTP2rcYfn6RC-Y/edit#bookmark=id.1ntpraugze7) we discussed the reduced participation in the [Unlabelled issues triage package](https://about.gitlab.com/handbook/engineering/quality/triage-operations/#newly-created-unlabelled-issues-requiring-first-triage) and would like feedback on how it can be improved. ### Improvements - Utilize OOO status indicator for triage package - https://gitlab.com/gitlab-org/quality/triage-ops/issues/162 - Investigate making PM DRI for check for duplicate issues - https://gitlab.com/gitlab-org/quality/triage-ops/-/issues/505 - Reduce load with https://gitlab.com/gitlab-org/quality/team-tasks/-/issues/295

2019-11-18 21:06:56.593

Apply ~documentation to Open ~"Community contribution" documentation MRs

### Proposal Apply ~documentation to open ~"Community contribution" MRs that have changes to `docs/`. As @godfat suggested, this might be easier than trying to react with Triage-Serverless. From https://gitlab.com/gitlab-org/technical-writing/issues/39#note_240604168 to help facilitate ~"Community contribution" doc improvements.

2019-11-04 07:37:50.078

Label missed-deliverable and missed milestones after the milestone expires but before the 22nd

This is due to the confusion of the EMs, our self-managed customers, the wider community and also a request from the delivery team. We have 2 dates to juggle for a given release. 1. On the 17th when the milestone expires. 1. When we release on the 22nd. Currently, there is still a possibility of things being part of the release even if the issue is closed after the 17th but before the 22nd. This is due to the time window between these 2 dates. There is confusion due to bot labeling. Some ~"missed-deliverable" are still being delivered since labeling on the 17th was too soon. After discussing with @marin the general idea is to leave the milestone as they are and tack on the additional buffer with the bot labelling. Until we can actually infer via automation, how every change makes it through to production, we have the below as an iteration. ### Proposal Label ~"missed-deliverable" and `missed:x.y` version on the last business day before the 22nd. * If the 22nd falls on a friday, we label this on thursday the 21th. * If the 22nd falls on a monday, we label this on friday the 19th. * If the 22nd falls on either weekend days, we label this on friday the 21th or the 20th. This labeling will happen on the last business day at midnight pacific time. #### Implementation For the logic, we are already pulling from the milestone dates. I am proposing that this should be an addon so we still rely on what is in `gitlab-org` milestones but calculate it in respect to the 22nd. * From a given milestone, find the last day (expire date) that will give us the month/year. * From the month/year look up what day is the 21st of that month/year. * If the 21st is a weekday, bot runs labelling on that day. * If the 21st is sunday, bot runs labelling on the 19th. * If the 21st is saturday, bot runs labelling on the 20th. Thoughts @godfat @gl-quality/eng-prod ? cc @marin @gitlab-org/delivery

2019-10-31 20:07:18.424

Setup Search team triage report and make Changzheng Liu, the new EM of Search, a recipient of `group::search` triage report

## Summary Search team is new. If the triage report has not been setup, please set it up in the bot. Please make @changzhengliu the recipient of this triage report. * [ ] Triage packages * Group triage packages * WIP: Infer stage and group labels from subject labels

2019-10-24 13:02:15.424

Adopt merge request pipelines

Following up from https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/325#note_235030694 We need to fully move to merge request pipelines, otherwise some pushes might create confusing two pipelines at the same time!  On the above screenshot, we can also see the pipelines are out of orders.

2019-10-24 07:36:34.614

Update `group::orchestration` triage package EM from @DylanGriffith to @sengelhard

2019-10-22 08:14:23.591

Add a way to automatically manage pipeline schedules

We start to have many pipeline schedules so that would be easier to have a script that automatically manages them for us, based on a specification YAML file. 1. Pipeline schedules are specified in a `pipeline-schedules.yml` YAML file 1. Pipeline schedules are synchronized automatically on every `master` commit 1. The same job runs with `--dry-run` on MRs

2019-09-27 08:12:45.429

Notification ping when the number of unique Zendesk or Salesforce links reach a threshold

The step 2 at https://gitlab.com/gitlab-org/quality/triage-ops/issues/156 * Count the number of unique zendesk links in the discussions / comments. * If 5 and above mention the Engineering Manager and Product Manager for that group. * If more than 7 and above mention the Dir of Product for that group. * No need to store any caching we just get a count before adding logic to the notification.

2019-09-20 22:16:56.160

Make priority and severity scoped labels

Per @tipyn in https://gitlab.slack.com/archives/C3JJET4Q6/p1568852155060400 We also have discussed making ~S1 ~S2 .. and ~P1 ~P2 labels scoped before. https://gitlab.com/gitlab-org/gitlab/issues/27906#priorities-and-severities Right now as teams and groups are getting better at using these labels we should consider changing them to scoped labels so that only one of the labels are present when changing. ### Priority * ~P1 to ~"priority::1" * ~P2 to ~"priority::2" * ~P3 to ~"priority::3" * ~P4 to ~"priority::4" ### Severity * ~S1 to ~"severity::1" * ~S2 to ~"severity::2" * ~S3 to ~"severity::3" * ~S4 to ~"severity::4" I believe the next steps are to estimate the work and changes need in all of our systems. Triage automation and charts to reflect this. ## Steps: 0. [x] `gitlab-org/gitlab` :arrow_right: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/39136 and https://gitlab.com/gitlab-org/gitlab/-/merge_requests/40280 0. [x] `gitlab-org/release-tools` :arrow_right: https://gitlab.com/gitlab-org/release-tools/-/merge_requests/1151 0. [x] `gitlab-com/www-gitlab-com` :arrow_right: https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_requests/59662 0. [x] `gitlab-org/quality/triage-ops` :arrow_right: https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/574 0. [x] `gitlab-org/gitlab-insights` :arrow_right: https://gitlab.com/gitlab-org/gitlab-insights/-/merge_requests/197 0. [x] `gitlab-org/quality/insights-config` :arrow_right: https://gitlab.com/gitlab-org/quality/insights-config/-/merge_requests/13 0. [x] `gitlab-com/gl-infra/incident-management` :arrow_right: https://ops.gitlab.net/gitlab-com/gl-infra/incident-management/-/merge_requests/18 0. [x] `gitlab-com/gl-infra/triage-ops` :arrow_right: https://gitlab.com/gitlab-com/gl-infra/triage-ops/-/merge_requests/43 0. [x] `gitlab-com/gl-infra/production` :arrow_right: https://gitlab.com/gitlab-com/gl-infra/production/-/merge_requests/34 0. [x] (merge around the time we rename) `gitlab-org/quality/triage-serverless` :arrow_right: https://gitlab.com/gitlab-org/quality/triage-serverless/-/merge_requests/29 0. [x] Rename the labels to scoped labels for `gitlab-org` https://gitlab.com/groups/gitlab-org/-/labels?search=P1 0. [x] Rename the labels to scoped labels for `gitlab-com` https://gitlab.com/groups/gitlab-com/-/labels?search=P1 0. [x] (after renamed) `gitlab-org/release-tools` :arrow_right: https://gitlab.com/gitlab-org/release-tools/-/merge_requests/1155 @gl-quality/eng-prod

2019-09-13 21:50:34.808

Add charts/gitlab to milestone/deliverable triage

## Summary This issue is for requesting triage automation on a given project. Please link to the project needed and select from a list of triage automation rules to be added. Project: https://gitlab.com/gitlab-org/charts/gitlab * Automation that does not require an assignee. * Rescheduling of milestones (milestone grooming) * Labelling ~"missed-deliverable" and `missed:x.y` based on the scheduled milestone. The Distribution team is slowly working on aligning our process to the rest of the gitlab teams. We aren't ready yet to opt-in to all daily tasks, but we have switched to using the `Deliverable` label, so can now make use of the missed deliverable handling. I've taken a look, and I think we can use the `move-milestone-forward.yml` and `labels-reminders.yml` policies.

2019-09-13 21:46:03.024

Add gitlab-omnibus to milestone/deliverable triage

## Summary This issue is for requesting triage automation on a given project. Please link to the project needed and select from a list of triage automation rules to be added. Project: https://gitlab.com/gitlab-org/omnibus-gitlab/ * Automation that does not require an assignee. * Rescheduling of milestones (milestone grooming) * Labelling ~"missed-deliverable" and `missed:x.y` based on the scheduled milestone. The Distribution team is slowly working on aligning our process to the rest of the gitlab teams. We aren't ready yet to opt-in to all daily tasks, but we have switched to using the `Deliverable` label, so can now make use of the missed deliverable handling. I've taken a look, and I think we can use the `move-milestone-forward.yml` and `labels-reminders.yml` policies.

2019-09-12 08:30:02.248

Rename `roadmaps` to `Category:Roadmaps`, `Category:Agile Portfolio Management` to `Category:Epics`

## Summary Once https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/29027/diffs is merged, rename `roadmaps` to `Category:Roadmaps`, and `Category:Agile Portfolio Management` to `Category:Epics`. * ~"roadmaps": [https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name\[\]=roadmaps](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name%5B%5D=roadmaps) * ~"Category:Agile Portfolio Management": [https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name\[\]=Category%3AAgile%2FPortfolio%2FManagement](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name%5B%5D=Category%3AAgile%2FPortfolio%2FManagement) For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at https://quality-dashboard.gitlap.com/groups/gitlab-org are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [ ] (If applicable) Dashboard creation: create or update the stage/group in https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb. * [ ] (If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests. * [ ] (If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name. * [ ] (If applicable) Delete the old dashboard views using the deprecated labels. * [ ] (If applicable) Update the group triage package definition to use the new label /cc @gl-quality/eng-prod @kokeefe @johnhope @donaldcook

2019-09-12 07:59:51.906

Rename `search` to `Category:Search`

## Summary Rename `search` to `Category:Search`. * ~"search": [https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=search](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name%5B%5D=search) For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [ ] (If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. * [ ] (If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests. * [ ] (If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name. * [ ] (If applicable) Delete the old dashboard views using the deprecated labels. * [ ] (If applicable) Update the group triage package definition to use the new label /cc @gl-quality/eng-prod @phikai @craig-gomes

2019-09-11 18:03:04.096

Growth groups will be renamed

## Summary Please list the Stage, Group or Category labels changes required. The change can fall under the 3 scenarios below: * Renaming an existing Stage or Group ~"group::activation" will be `group::acquisition` * Renaming an existing Stage or Group ~"group::upsell" will be `group::conversion` * Renaming an existing Stage or Group ~"group::adoption" will be `group::expansion` For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [ ] (If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. => https://gitlab.com/gitlab-org/gitlab-insights/merge_requests/163 * [x] ~~(If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests.~~ * [x] ~~(If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name.~~ * [x] Rename the group labels. * [x] ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~ * [ ] (If applicable) Update the group triage package definition to use the new label => https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/304

2019-09-11 16:16:35.259

Dockerfile.dev mismatch with production

`Dockerfile.dev` was using https://gitlab.com/Alexand/ruby-function-invoker but now it seems production is using https://gitlab.com/gitlab-org/serverless/runtimes/ruby which behaves differently. We need to bring them align.

2019-09-02 08:46:07.826

Add exception so the prep/backport MRs for patch releases never have their milestones updated

The milestone was updated by the bot on https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/15885 ``` 🤖 GitLab Bot 🤖 @gitlab-bot changed milestone to %12.3 8 hours ago 🤖 GitLab Bot 🤖 @gitlab-bot added missed:12.2 label 8 hours ago ``` Because these are backport MRs the milestone should stay 12.2, perhaps we should add a new label to identify these MRs since they are special. We also, potentially have backport MRs that developers open against 12.2 that will be targeting the `12-2-stable[-ee]` branches. If we could skip anything targeting `MAJOR-MINOR-stable*` that would suffice I think.

2019-08-27 08:13:23.266

`VersionedMilestone#group_milestone` incorrect milestone query when source_id is not a path

Job [#280629926](https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/280629926) failed for c7e8a8da969a3518bf8895b8af800647e2f74284: Related to changing to using numeric IDs for projects and groups across our rules

2019-08-26 07:29:00.628

Do not fail the schedules if tests won't pass

Now we have two incidents due to changes from www-gitlab-com: * https://gitlab.com/gitlab-org/quality/triage-ops/issues/253 * https://gitlab.com/gitlab-org/quality/triage-ops/issues/287 I think we need to do something here. Here are a few options: * Use `webmock` to avoid dependencies on www-gitlab-com * Remove those backward compatible tests: https://gitlab.com/gitlab-org/quality/triage-ops/issues/253#note_201456403 * Avoid running tests on schedules * Make tests allow to fail on schedules @gl-quality/eng-prod Which would you prefer?

2019-08-23 06:01:07.463

Label `missed:x.y` as soon as the milestone expires and not when we close and reschedule issues and MRs

We currently label ~"missed-deliverable" when a milestone expires. However we do not label them with `~missed:x.y` e.g. ~"missed:12.2" and etc. We only do this after the milestone is closed. This can cause a short time where an issue detected as ~"missed-deliverable" is re-scheduled by a human manually into the next milestone. This would result in an issue with ~"missed-deliverable" but we cannot pin down which version it was missed. This is likely the cause of the grey undefined missed deliverables. For example see https://gitlab.com/gitlab-org/gitlab-ce/issues/59754#note_207090158, I had to add ~"missed:12.2" manually. This will get labeled tomorrow since %12.2 is now closed.  ## Proposal * Let's label both ~"missed-deliverable" and `missed:x.y` labels together as soon as the milestone expires. @kwiebers or @markglenfletcher maybe one of you can help take this up? I would like someone other than @godfat to do it to spread the knowledge here. Please see work that @godfat has done prior, this should be an easy tweak to the triage rules. https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/163/diffs Also @joshlambert for awareness.

2019-08-22 12:04:23.187

Consider creating a CI config file per project

We should consider creating one CI config file per project and explicitly list the jobs that should run for each project. That would clarify what are the policies that run for each project.

2019-08-20 16:45:51.771

Rename ~"group::serverless and paas" label to ~"group::serverless"

## Summary Following https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/28665, we will need to rename of ~"group::serverless and paas" label to ~"group::serverless". For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [x] (If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. => https://gitlab.com/gitlab-org/gitlab-insights/merge_requests/155 * [ ] ~~(If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests.~~ * [ ] ~~(If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name.~~ * [ ] ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~

2019-08-20 16:45:28.992

Rename ~"group::team planning" to ~"group::project management"

## Summary Following https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/28178/diffs, we will need to rename of ~"group::team planning" to ~"group::project management". For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [x] (If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. * [x] (If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests. * [x] (If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name. * [x] (If applicable) Delete the old dashboard views using the deprecated labels. /cc @gl-quality/eng-prod

2019-08-20 16:43:50.584

Fix stage label detection to disallow nested scoped labels

It was brought to my attention that some nested scoped labels starts as a stage label, but then have an additional scoping, e.g. `~devops::release::feature flags` (currently only exists for the `gitlab-ee` project). We should fix the logic to only detect stage labels matching something like `/\Adevops::\w+\z/`.

2019-08-20 10:15:27.562

Stage/group labels inference doesn't work for confidential issues

Due to how `gitlab-triage` works by default, it redact attributes of confidential resources, so we cannot infer labels in the `actions.comment` action since attributes are redacted. We can work-around that by setting `@redact_confidentials = false` before performing the inference for now.

2019-08-20 09:54:35.772

Consider using group/project IDs instead of path in jobs and pipeline schedules for better stability

The following discussion from !243 should be addressed: - [ ] @zj-gitlab started a [discussion](https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/243#note_205556529): (+1 comment) > Bit of a nitpick, but I think using project IDs are more stable. For example, I've changed my username to `zj-gitlab` and the old username still gets mentions from bots that work by username instead of ID. Not sure how often we move projects, but could be a potential issue. I think that's a good suggestion, and I think it would be fine since the project/group name should be present in the job or pipeline schedule. I'll resolve that in a separate issue, thanks! This is also relevant with the single codebase changes approaching.

2019-08-19 12:25:12.051

Consider cleaning up Ecosystem-specific labels

It seems the Ecosystem group has a lot of labels, even duplicate labels. ## `gitlab-org` Group labels - ~"group::ecosystem::integrations" - ~"group::ecosystem::marketplace" - ~"group::ecosystem::sdk" These are categories labels and should follow the `Category::<category name>` convention documented at https://about.gitlab.com/handbook/marketing/website/#category-attributes under `label`. ## `gitlab-org/gitlab-ce` Project labels - ~"ecosystem::integration" - ~"ecosystem::integration::chat" - ~"ecosystem::integration::cicd" - ~"ecosystem::integration::infra" - ~"ecosystem::integration::jenkins" - ~"ecosystem::integration::metrics" - ~"ecosystem::integration::other" - ~"ecosystem::integration::security" - ~"ecosystem::integration::workflow" - ~"ecosystem::integration::marketplace" - ~"ecosystem::marketplace" duplicates the `gitlab-ce` ~"ecosystem::integration::marketplace" and the `gitlab-org` ~"group::ecosystem::marketplace" - ~"ecosystem::sdk" duplicates the `gitlab-org` ~"group::ecosystem::sdk" I'm not in favor of triple-nested scoped labels, and these shouldn't be project-specific labels, but group labels instead (we could just promote them to group labels).

2019-08-18 20:34:04.944

Remove the ~"group::analytics" & ~"devops::manage" labels when they're associated with the ~project, ~"gitlab.com" or ~navigation labels

Hi, Can we please move the issues with tags `project`, `gitlab.com` and `navigation` to `group::unallocated`? I think it makes sense to have them with a tag, so that management is aware how many of those are without anyone looking after them (https://gitlab.slack.com/archives/C0NFPSFA8/p1565991373232800). cc: @Jeremy, @ebrinkman This issue is a follow up of: https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/27707#note_203529447 and the series thereof.

2019-08-16 16:23:04.338

Improve community contribution MR discovery and coaching

This was brought up organically by @joshlambert that some of the MRs for ~"devops::verify" from the community are building up. Some of them being almost 1 year. > Josh: We have a huge backlog of community contributed MR’s in Verify (specifically Runner) of over 100 (~20% of total). This is proving to be a very frustrating experience, not just contributors but also paying customers: https://gitlab.com/gitlab-org/gitlab-runner/merge_requests/1041#note_202338266. => Josh: Mek, can we add the backlog of community MR’s to triagebot and quality dashboard? => Mek: We already have a Community Contribution triage package, I‘ll follow up on making improving the format. We have community MR triage packages for MRs already https://about.gitlab.com/handbook/engineering/quality/triage-operations/index.html#community-merge-requests-requiring-attention This seems to only be active for `GitLab CE` & `GitLab EE` I think we should expand this to other core projects e.g. * [x] Gitaly - https://gitlab.com/gitlab-org/quality/triage-ops/pipeline_schedules/29055/edit * [x] Runner - https://gitlab.com/gitlab-org/quality/triage-ops/pipeline_schedules/29680/edit Other improvements: * [x] Consider adding ~"Community contribution" to this type of triage report so we can filter on ~"triage-package" and ~"Community contribution" => !260 * https://gitlab.com/gitlab-org/gitlab-ce/issues/58131 * [x] Spell out the names of these triage reports. Instead of `Merge requests requiring attention` call it `Community Contributions MR report` => !260 * [x] Add a daily report for untriaged 'Community contribution' MRs: https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/592 * [x] Draw attention to EMs for ~"Community contribution" based on a `~group::` label: https://gitlab.com/gitlab-org/quality/triage-ops/-/issues/565 => https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/622 * [ ] Post these Community Triage package on the #mr-coach channel

2019-08-14 12:09:04.312

New policy to move all open issues from `gitlab-org/gitlab-foss` to `gitlab-org/gitlab`

As part of https://gitlab.com/gitlab-org/gitlab-ee/issues/13304, `gitlab-org/gitlab-foss` issue tracker will stay open but we want to automatically move new issues to `gitlab-org/gitlab`. We should leave an explanation message when doing so, and this should set up as a daily hygiene automation (or even a serverless automation).

2019-08-13 08:08:28.675

Who's the correct PM for the Runner group?

In https://gitlab.com/gitlab-org/quality/triage-ops/blob/a730c493e301bd052e8f54142540ece57d216553/policies/team-triage-package.yml#L1065-1066 the PM for the Runner group is set to @jlenny but in https://about.gitlab.com/handbook/product/categories/#runner-group it's @joshlambert? What is the correct PM for ~"group::runner"?

2019-08-13 08:03:09.196

Create a triage package for the Import group

I just noticed that the [Import group](https://about.gitlab.com/handbook/product/categories/#import-group) doesn't have a triage package!

2019-08-12 23:52:21.203

Software Composition Analysis group will be renamed to Composition Analysis

## Summary From https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/26742, ~"group::software composition analysis" will be renamed to ~"group::composition analysis" For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [x] Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. => https://gitlab.com/gitlab-org/gitlab-insights/merge_requests/150 * [x] No need to migrate anything since the label will be renamed. ~~(If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests.~~ * [x] ~~(If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name.~~ * [x] Same here, the existing dashboard will just be changed to use the new label name. ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~ /cc @gl-quality/eng-prod

2019-08-12 23:19:09.514

Update Larissa Lane as PM for group::distribution and group::memory

We have a new PM for ~"group::distribution" and ~"group::memory". Please add ownership from product in all triage automation to `Larissa Lane` @ljlane. @gl-quality/eng-prod /cc @ebrinkman

2019-08-08 22:50:15.038

Search group will be part of enablement

## Summary ~group::search will be created under ~"devops::enablement" https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/27493#note_201799245 All search features and `~category::` will be moved here as well. For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [x] Triage automation: create or update the stage/group/categories in <https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/lib/devops_labels.rb>. * [x] (If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. - https://gitlab.com/gitlab-org/gitlab-insights/merge_requests/149 * [ ] (If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests. - https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/246 * [ ] ~~(If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name.~~ * [ ] ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~ /cc @phikai /cc @gl-quality/eng-prod

2019-08-08 17:55:58.717

Default Group Label for Specific Projects

## Proposal Add new policies for each project that will be scheduled and run prior to the daily label inference policy. ## Background and Research @phikai reported that Issues that were labeled with ~search had the ~"devops::create" and ~"group::editor" labels applied by GitLab Bot. The impacted issues look to be issues without a group label present and have a matching category label which will be used to infer the stage and group label. Examples looking at ~search : - https://gitlab.com/gitlab-org/gitter/webapp/issues/1510 - https://gitlab.com/gitlab-org/gitlab-docs/issues/167 - https://gitlab.com/gitlab-org/gitter/webapp/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=search - https://gitlab.com/gitlab-org/gitlab-docs/issues?scope=all&utf8=%E2%9C%93&state=opened&label_name[]=search While I was looking at the examples in the searches, it seemed that there was not a group label on the impacted issues. On an ongoing basis, there will be a comment informing why the labels were inferred but we disabled that for the bulk runs in https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/199/pipelines @markglenfletcher had the idea to default group labels for certain projects. Here's some of the mappings from the groups he identified if we want to go that route. - [Gitaly](https://gitlab.com/gitlab-org/gitaly) - ~"group::gitaly" - [Runner](https://gitlab.com/gitlab-org/gitlab-runner) - ~"group::runner" - [Gitter](https://gitlab.com/gitlab-org/gitter) - ~"group::gitter" - [Omnibus](https://gitlab.com/gitlab-org/omnibus-gitlab) - ~"group::distribution" - [GitLab CNG](https://gitlab.com/gitlab-org/build/CNG) - ~"group::distribution" There is likely more that I'm missing but I started with what I could find and felt confident in :smile:

2019-08-08 14:15:30.008

Deprecate the legacy team labels

## Summary We will deprecate the following legacy team labels according to the new rollout plan at https://gitlab.com/gitlab-org/quality/triage-ops/issues/172#rollout-steps by adding a ` [DEPRECATED]` suffix to them: - [x] ~Manage - [x] ~Plan - [x] ~Create - [x] ~Verify - [x] ~Package - [x] ~Release - [x] ~Configure - [x] ~Serverless - [x] ~Monitor - [x] ~Secure - [x] ~Defend - [x] ~Growth - [x] ~Gitaly - [x] ~Gitter - [x] ~Distribution - [x] ~Geo - [x] ~Memory - [x] ~Ecosystem Groups should be using exclusively their group and stage labels from now on. ## Action items ### On 2019-08-13 - [x] Announce the deprecation on the company call, `#development`, `#product`, and Engineering Week in Review. >>> On 2019-08-19, we will deprecate the team labels by adding a ` [DEPRECATED]` suffix to them (https://gitlab.com/gitlab-org/quality/triage-ops/issues/254). Please use the Stage (`devops::xxx`) and Group (`group::xxx`) labels instead. - This is part of an effort to transition from team labels to stage & group labels: https://gitlab.com/gitlab-org/quality/triage-ops/issues/172#note_201666990 - We are tackling the remaining edge cases where more than one legacy team labels exists after this https://gitlab.com/gitlab-org/quality/triage-ops/issues/242. >>> ### Before 2019-08-19 - [x] Triage automation: create or update the stage/group/categories in <https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/lib/devops_labels.rb>. => https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/255 - [x] Dashboard creation: update the teams in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. => https://gitlab.com/gitlab-org/gitlab-insights/merge_requests/151 - [x] ~~(If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests.~~ - [x] The native Insights configuration - [x] The Periscope dashboards - [x] Update https://gitlab.com/gitlab-org/gitlab-ce/blob/master/doc/development/contributing/issue_workflow.md => https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/31811 ### On 2019-08-19 - [x] Archive the old labels with renaming and adding "DEPRECATED" at the end of the label names. - [x] ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~

2019-08-07 16:35:29.668

Replace blocked label with workflow::blocked

## Summary To assist in reporting and reduce duplicate labels, we will be conslidating ~"blocked [deprecated in favor of workflow : : blocked]" and ~"workflow::blocked". To do this, we will need to rename both labels to reduce the amount of change to [Org Boards]() and email notifications triggered by GitLab Bot. Currently the ~"workflow::blocked" label is used less than the ~"blocked [deprecated in favor of workflow : : blocked]" so we will rename ~"blocked [deprecated in favor of workflow : : blocked]" to ~"workflow::blocked" and migrate the existing ~"workflow::blocked" to the new label. | Type of Resource | ~"workflow::blocked" Count | ~"blocked [deprecated in favor of workflow : : blocked]" Count | | ------ | ------ | ------ | | Boards | 10 | 16 | | Issues | [23](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=%E2%9C%93&state=all&label_name[]=workflow%3A%3Ablocked) | [164](https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=all&label_name[]=blocked%20%5Bdeprecated%20in%20favor%20of%20workflow%20%3A%20%3A%20blocked%5D) | | MRs | [3](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=✓&state=all&label_name[]=workflow%3A%3Ablocked) | [63](https://gitlab.com/groups/gitlab-org/-/merge_requests?scope=all&utf8=%E2%9C%93&state=all&label_name[]=blocked%20%5Bdeprecated%20in%20favor%20of%20workflow%20%3A%20%3A%20blocked%5D) | Boards were counted by looking at group boards at the gitlab-org level: https://gitlab.com/api/v4/groups/gitlab-org/boards ### Board List - [In dev / In review](https://gitlab.com/groups/gitlab-org/-/boards/364214) - [Health - Workflow](https://gitlab.com/groups/gitlab-org/-/boards/1160198) - [Geo Kanban Build](https://gitlab.com/groups/gitlab-org/-/boards/1181257) - [Monitor kick off - Workflow](https://gitlab.com/groups/gitlab-org/-/boards/1223141) - [Plan Kanban](https://gitlab.com/groups/gitlab-org/-/boards/1226305) - [12.3 Package](https://gitlab.com/groups/gitlab-org/-/boards/1229228) - [Team Planning Kanban](https://gitlab.com/groups/gitlab-org/-/boards/1235826) - [Certify Kanban](https://gitlab.com/groups/gitlab-org/-/boards/1235846) - [Miranda Fluharty](https://gitlab.com/groups/gitlab-org/-/boards/1241526) - [Manage::Analytics::Product Process](https://gitlab.com/groups/gitlab-org/-/boards/1244153) ## Action items * [ ] ~~Triage automation: create or update the stage/group/categories in <https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/lib/devops_labels.rb>.~~ * [ ] ~~(If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>.~~ * [ ] (If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests. => https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/238/diffs * [x] (If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name. * [ ] ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~ * [ ] Rename the existing labels based on the following rules - [ ] `~workflow::blocked` -> `~workflow::blocked-migrate` - temporary to free up scoped label - [ ] `~blocked [deprecated in favor of workflow : : blocked]` -> `~"workflow::blocked" * [ ] Announce that ~"blocked (deprecated) — use workflow : : blocked instead" will be deleted in favor of ~"workflow::blocked" * [ ] Company call - Communicated on 8/14 * [ ] `#development` * [ ] `#product` * [ ] Delete the ~"blocked [deprecated in favor of workflow : : blocked]" label /cc @gl-quality/eng-prod

2019-08-06 11:29:48.314

A breaking API change broke the `policies/label-accepting-merge-requests.yml` policy

A breaking change to the Issues and MRs API was introduced and reported in https://gitlab.com/gitlab-org/gitlab-ce/issues/65676. This in turn made our `policies/label-accepting-merge-requests.yml` policy apply the ~"Accepting merge requests" to a lot of issues that don't have a milestone, where normally they shouldn't (as we target issues with a milestone only). A fix is in preparation to prevent that from happening while the API issue is fixed and deployed, and to revert the mistakenly added labels: https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/233

2019-08-05 22:27:48.492

Consolidate more labels for marking issues that are open for community contributions

### Problem Following the changes in https://gitlab.com/gitlab-org/quality/triage-ops/issues/179, there are three more label variants for marking issues that are open for community contributions. These include: 1. **7 issues** with ~"Accepting Merge Requests" project label used in the [`gitlab-workhorse`](https://gitlab.com/gitlab-org/gitlab-workhorse) project. 2. **2 issues** with ~"accepting merge requests" project label used in the [`gitlab-mattermost`](https://gitlab.com/gitlab-org/gitlab-mattermost) project. 3. **3 issues** with ~"accepting MR" project label used in the [`cookbook-gitlab`](https://gitlab.com/gitlab-org/cookbook-gitlab) project. The issues affected are very few but this can cause some confusion during issue search.  ### Proposal Let's remove these three project labels and replace them on all attached issues with the group label ~"Accepting merge requests".

2019-08-01 03:47:02.917

Remove association between `gitlab.com` label and Manage/Analytics

## Overview GitLab Bot appears to apply ~"devops::manage" and ~"group::analytics" (or possibly Measure) to issues associated with the ~"gitlab.com" label. This isn't as helpful as we'd like, since ~"gitlab.com" issues span more than just Manage issues - see https://gitlab.com/gitlab-org/gitlab-ce/issues/62052 for an example. ## Proposal Either: * Remove the association for `gitlab.com` and don't automatically apply a stage/group label, or * Don't attempt to reapply stage/group labels for `gitlab.com` issues. cc @meks @rymai @markglenfletcher

2019-08-01 03:07:24.320

Un-scope category labels

### Problem to solve The label documentation mentions `Subject` labels: https://gitlab.com/gitlab-org/gitlab-ce/blob/master/doc/development/contributing/issue_workflow.md#subject-labels but it currently mixes feature and category labels. We need to clarify this. For the distinction between **Category** and **Feature**, see https://about.gitlab.com/handbook/product/categories/#hierarchy There is no official naming/color convention for these labels but an ongoing normalization proposal with https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/169 suggest to leverage scoped labels. I don't think scoped labels are useful here and they even prevent some existing workflow (e.g. having an [epic related to multiple categories](https://gitlab.com/groups/gitlab-org/-/epics/971)). Requiring to split into sub-epics/issues will generate unnecessary overhead. NB: There is a dedicated proposal to normalize feature labels: https://gitlab.com/gitlab-org/quality/triage-ops/issues/247, this issue focuses on feature labels. ### Example Issues/Epics are not only about product/category strategies, but there are also engineering initiatives and some of them cover multiples categories depending on how the code is structured. E.g. in ~"devops::secure" we have a lot of overlapping areas between the groups and categories as the code is generic. One unique change to the code can impact several categories. This MR https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/14376 impacts 4 categories. **Without scoped labels:** - we can have 1 epic, 1 issue, 1 MR, all labeled with the 4 category labels. Simple and straightforward. **With scoped labels:** - we need 1 parent epic, 4 child epics, 4 issues, all labeled independently with 1 category label. - the MR can't have the category labels reflecting the change it's providing. - these are way too much useless items to maintain. Engineers need to update `workflow::xxx` labels and close issues after deployed and tested on prod. This also clutters the issue boards for no benefit. ### Proposal: - [x] Naming convention `Category: Feature with Spaces And Capital Letters as Desired` - [x] No scoped labels (mutual exclusion). - [x] Color `#428BCA` Once we agree on the convention, tasks list: - [x] Update the docs by creating a distinct `Category labels` section => https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/32053/diffs - [x] Update any label references in `www-gitlab-com` => https://gitlab.com/gitlab-com/www-gitlab-com/merge_requests/28532/diffs - [x] Update the scripts/automation in this project => https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/274/diffs - [x] Once the three above MR are merged, update the labels according to the naming convention and color - [x] Communicate broadly - [x] Company call - [x] Slack: `#development`, `#quality`, `#product`

2019-07-31 15:50:27.207

Add group and stage label inference to GitLab Charts

The Helm Charts are stored within the GitLab [charts](https://gitlab.com/charts) group and not on gitlab-org. This means label inference is not run on the charts projects. Applying the labels will enable issue and merge requests reporting by group/stage for the Charts projects. This would cause some chatter on issues & merge requests from GitLab Bot. Charts are getting moved into gitlab-org eventually: https://gitlab.com/gitlab-org/distribution/team-tasks/issues/422 ## Dry Run Results Pipeline runs (as of 8/5): - `stage-and-group-labels-hygiene.yml` - 69 matching issues: https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/266437366 - `mr-migrate-legacy-labels.yml` - 15 open merge requests & 294 merged merge requests: https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/266438245 Looking for feedback from @gitlab-org/distribution

2019-07-30 22:19:29.521

New group::import is now in effect for devops::manage

## Summary Please list the Stage, Group or Category labels changes required. The change can fall under the 3 scenarios below: * Adding a new Stage or Group - ~"group::import" under ~"devops::manage" For every case above, the Engineering Productivity team needs to ensure that: * The label change is factored into the triage mechanism. * Engineering Dashboards at <https://quality-dashboard.gitlap.com/groups/gitlab-org> are updated. * Old labels are migrated correctly on affected issues and merge requests. ## Action items * [x] Triage automation: create or update the stage/group/categories in <https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/lib/devops_labels.rb>. => https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/215 * [x] (If applicable) Dashboard creation: create or update the stage/group in <https://gitlab.com/gitlab-org/gitlab-insights/blob/master/lib/gitlab_insights.rb>. => https://gitlab.com/gitlab-org/gitlab-insights/merge_requests/140 * [x] (If applicable) Label migration on existing issues and merge requests: apply the new label to opened & closed issues, and open & merged merge requests. * [ ] ~~(If applicable) Archive the old label with renaming and adding "DEPRECATED" at the end of the label name.~~ * [ ] ~~(If applicable) Delete the old dashboard views using the deprecated labels.~~ /cc @gl-quality/eng-prod @jeremy

2019-07-30 19:08:05.582

Remove missed-SLO label from all P3 and P4 Bugs

Identify all P3 and P4 bugs which have a ~"missed-SLO" label and remove it. These are feeding into insights reporting that needs to be removed per https://gitlab.com/gitlab-com/Product/issues/376 @meks @rymai - I tweaked this a bit from our discussion this morning so that it's only looked at GitLab bot applied labels. What do you think?

2019-07-30 17:10:21.801

Proposal: change type labels to be scoped and create children labels per type

## Proposal 1. Change type labels to be scoped, e.g. `type::bug`, `type::backstage` 1. Create children scoped labels per type, e.g. `bug::ui`, `bug::performance`, `backstage::technical debt`, `backstage::static analysis`, `backstage::ci config` etc. This proposal is extracted from the discussion at https://gitlab.com/gitlab-org/quality/triage-ops/issues/88.

2019-07-30 14:16:23.200

Re-enable the "intelligent" stage/group label inference for open issues

Once https://gitlab.com/gitlab-org/quality/triage-ops/issues/234 is done, all closed and open issues should have a stage/group label set from their team label (except for the ones with multiple team labels), so we should be able to re-enable the "intelligent" stage/group label inference **for open issues**, in the daily schedule.

2019-07-30 14:14:37.093

Perform a 1:1 team -> stage/group inference for closed issues and merge requests

As discussed in today's Engineering Productivity call, we should perform the 1:1 team -> stage/group inference for closed issues and merge requests as well, for historical data consistency.

2019-07-29 14:20:38.562

Add @gitter-badger to the `gitlab-org` group

### Problem Following the problem described in https://gitlab.com/gitlab-org/quality/triage-ops/issues/174, contributions by @gitter-badger are also marked with the ~"Community contribution" label. See [merge requests](https://gitlab.com/gitlab-org/gitter/webapp/merge_requests?scope=all&utf8=%E2%9C%93&state=opened&author_username=gitter-badger) in `gitlab-org/gitter` group. ### Proposal Following the relevant slack [discussion](https://gitlab.slack.com/archives/C3W3PSR88/p1564361438009100) (internal), let's move the @gitter-badger as a member (_Developer_) in the `gitlab-org` group. The description of the bot has been updated to include purpose and ownership.

2019-07-26 18:00:56.105

Temporarily remove missed SLO detection for P3

We currently have a high number of ~P3 backlog. We need to allow more flexibility and headroom for Product to groom the existing P3 bugs. As such, we have decided to temporarily disable the missed SLO detection for ~P3 for the time being and focus on ~P1 and ~P2 @godfat @rymai @markglenfletcher /cc @joshlambert @clefelhocz1 @craig-gomes @tpazitny

2019-07-26 16:37:56.313

Rename RUN_DAILY and RUN_WEEKLY

On !195, we found that the naming of the variables for `RUN_DAILY` and `RUN_WEEKLY` were named in an unclear manner. These variables are important to help group jobs into a collection to run within a single pipeline execution. From what I can tell, here's what what policies are run behind those two variables. Everything else seems to run in a `RUN_SINGLE` pipeline. ## Run daily - [Label reminders](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/labels-reminders.yml) - [Stage and group label hygiene](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/stage-and-group-labels-hygiene.yml) - [MR migrate legacy labels](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/mr-migrate-legacy-labels.yml) - [Accepting merge requests](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/label-accepting-merge-requests.yml) - [Move milestone forward](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/move-milestone-forward.yml) - [Discover](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/discover.yml) - [Missed SLO](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/label-missed-slo.yml) - [Unlabelled issues](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/unlabelled-issues.yml) - only when `UNLABELLED_TRIAGE_PACKAGE` is also set ## Run weekly - [Team triage package](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/team-triage-package.yml) - [Missing categories](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/missing-categories.yml) - [Community merge requests](https://gitlab.com/gitlab-org/quality/triage-ops/blob/master/policies/community-merge-requests.yml) - only when $COMMUNITY_TRIAGE_PACKAGE is also set ## Proposal There are currently two different types of triage policies that I group all policies into. - `TRIAGE_PACKAGE` - classification of incoming issues and merge requests for a group or project to aid in identifying actionable items for Product and Group/Stage Managers. - `TRIAGE_HYGEINE` - label or comment automation to aid in reporting or nudging behavior into desired direction. The daily jobs that don't have `packages` in the job name fit this category to me. The job name would be where we can indicate the desired frequency (`schedule:daily:reminders:labels` or `schedule:weekly:packages:community-merge-requests` ### Steps 1. Create new anchors based on the triage categories (ie `triage-package` and `triage-hygeine`) which leverage a different variable. 1. Convert existing jobs to use the new anchors 1. Update the README with the new variables description 1. After merge is completed, update the existing scheduled pipelines variables I am still getting my bearings and lack the historical context. I don't want to create a change that would be confusing. I also don't have great view of where Triage Packages 2.0 and 3.0 and how the evolution of our policies fit within these categories. @rymai @godfat @markglenfletcher - What are your thoughts? /cc @meks

2019-07-25 09:59:35.201

Split "CI & Runner" group

## Summary Splitting one "CI & Runner" group into two "Continuous Integration" and "Runner" groups. Please list the Stage or Group labels changes required. The change can fall under the 3 scenarios below: * Splitting a Stage into multiple Groups * Changing the Categories in a Stage * Adding a new Stage For every case above, the Engineering Productivity needs to ensure that: * The label change is factored into the triage mechanism. * Old labels are migrated correctly on affected Issues and Merge Requests. * Engineering Dashboards in https://quality-dashboard.gitlap.com/groups/gitlab-org is setup correctly as well. ## Action items * [x] Label migration on existing Issues and Merge Requests: apply the new label to closed/merged and in-progress work. * [x] Dashboard creation: create new dashboard view for new stage/group categories https://quality-dashboard.gitlap.com/groups/gitlab-org, location `gitlab_insights.rb` * [x] If applicable, archive the old labels with renaming and adding "DEPRECATED" at the end of the label name. * [x] If applicable, delete the old dashboard views using the deprecated labels. /cc @godfat @markglenfletcher @rymai @meks

2019-07-25 08:21:59.921

Update the triage operations guidelines to make sure that a heads-up is given in #development/#product/company call when an automation is expected to triage more than 100 issues/MRs at once

Following the recent 1:1 team -> stage/group inference automation (https://gitlab.com/gitlab-org/quality/triage-ops/issues/201), @gitlab-bot added stage and group labels to all the merge MRs under the `gitlab-org` group that had a single team label. We should give a heads-up in several Slack channels as well as the company call when we anticipate this amount of updates. We should document that in https://about.gitlab.com/handbook/engineering/quality/guidelines/triage-operations/.

2019-07-25 03:58:00.692

Milestone cleanup report renaming cuts off what was rescheduled

https://gitlab.com/gitlab-org/gitlab-ee/issues/13044 Issue was named as "Automatically rescheduled issues" but was renamed to "Automatically rescheduled to 12.2" the last word was truncated `issues` 

2019-07-24 15:27:00.544

Change the "intelligent" stage/group label inference logic to do the team -> stage/group inference first, and add a new "team to stage/group" inference logic

Change the "intelligent" stage/group label inference logic to: 1. Only use the 1:1 team -> stage/group inference for issues for now 1. Remove the 300 threshold for issues so that all open issues gets a stage/group from their single team label 1. Finish 1:1 team -> stage/group inference for issues 1. (Optional depending on the anticipated number of issues that would be updated) Re-add a threshold for issues 1. Use the "intelligent" inference for issues

2019-07-23 13:36:10.418

YAML lint in the pipeline?

To facilitate changes from outside the project we could lint the YML files for policies

2019-07-16 07:03:17.026

Close the merge request milestone movement report sooner

The report looks like: https://gitlab.com/gitlab-org/gitaly/issues/1784 We want to close it automatically, but because of the ordering we're processing the rules, this will only get closed in the next run. We should do better, but for now we can just wait for tomorrow. >>> The report was an issue, not a merge request, so this of course won't do anything. It doesn't close the issue right now because the report was only generated after we processed the merge requests. Whenever we're trying to close the report, the report for merge requests wasn't generated yet! So we need to run the same issue rule again after merge request rule is done. Well, or wait for tomorrow... >>> ---- The following discussion from !178 should be addressed: - [ ] @godfat started a [discussion](https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/178#note_189649103): (+2 comments) > Close for merge request report, too: https://gitlab.com/gitlab-org/gitlab-ce/issues/64291

2019-07-15 13:30:13.549

Make it possible to configure policies/unlabelled-issues.yml

Currently `policies/unlabelled-issues.yml` is a triaging package which has fixed names and people to rotate, so we can't reuse it for other groups. However we want to make it configurable so that we can easily apply to other groups. The current rotation code is: ``` ruby potential_triagers = %w[@markglenfletcher @godfat @ddavison @mlapierre @at.ramya @sliaquat @tnikic @zeffmorgan @tpazitny @wlsf82 @dchevalier2 @asoborov @grantyoung @jennielouie].shuffle list_items = resource[:items].split("\n") items_per_triagers = potential_triagers .zip(list_items.each_slice((list_items.size.to_f / potential_triagers.size).ceil)) .to_h.compact items_per_triagers.each_with_object([]) do |(triager, items), text| text << "#{triager}\n\n#{items.join("\n")}" end.join("\n\n") ``` And: ``` /assign #{items_per_triagers.keys.join(' ')} /label ~Quality ~"triage\-package" ``` We can aim to make it look like this: ``` ruby items_per_triagers = pick_triagers(resource[:items], ENV['TRIAGERS']) ``` And use it like: ``` ruby #{ items_for_triagers(items_per_triagers) } /assign #{items_per_triagers.keys.join(' ')} /label ~"#{ENV['TRIAGERS_GROUP']}" ~"triage\-package" ``` The following discussion from !183 should be addressed: - [ ] @markglenfletcher started a [discussion](https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/183#note_191755695): (+1 comment) > Approach looks good to me. I agree with letting these parts of the flow be disabled for now. > > I think that we should strive to enable the unlabelled/untriaged triage package jobs for all projects in the future, but instead of looping in the Quality team, allow a list of triage team members to be defined that are best suited to that project. I feel that this will allow teams to keep track of reported issues and increase responsiveness when communicating with issue authors.

2019-07-09 17:11:58.535

Extend daily/weekly schedules to other projects

Try to find the right balance for repeating and flexibility * [x] Gitaly: https://gitlab.com/gitlab-org/quality/triage-ops/issues/184 * [x] Runner: https://gitlab.com/gitlab-org/quality/triage-ops/issues/68 and https://gitlab.com/gitlab-org/quality/triage-ops/issues/164 * [x] QA

2019-07-08 20:58:10.698

Automation to ensure that issues and MRs with legacy team labels have a 1:1 mapping to their devops stage or group label

We need an on-going automated rule to apply `devops::xxxx` to issues and MRs based on their legacy team labels. This is the first step to ensure we can transition each groups to use the new labels. We would want a rule that continuously runs and auto adds the new stage and group labels where these old team labels exist. **Stages** * ~Manage => ~"devops::manage" * ~Create => ~"devops::create" * ~Plan => ~"devops::plan" * ~Verify => ~"devops::verify" * ~Package => ~"devops::package" * ~Release => ~"devops::release" * ~Configure => ~"devops::configure" * ~Monitor => ~"devops::monitor" * ~Secure => ~"devops::secure" * ~Defend => ~"devops::defend" * ~Growth => ~"devops::growth" * ~Enablement => ~"devops::enablement" **Groups** * ~Gitaly => ~"group::gitaly" * ~Gitter => ~"group::gitter" * ~Health => ~"group::health" * ~Distribution => ~"group::distribution" * ~Geo => ~"group::geo" * ~Memory => ~"group::memory" * ~Ecosystem => ~"group::ecosystem" #### Phases ##### Merge Requests 1. Add devops stage labels and (where necessary) group labels to merged Merge Requests that have a clear 1:1 mapping between legacy and stage label !184 - This accounts for the majority of throughputs and will make them visible in the stages dashboards - https://quality-dashboard.gitlap.com/groups/gitlab-org/stages/devops::create - https://quality-dashboard.gitlap.com/groups/gitlab-org/teams/create 2. Determine devops stage label to apply to merged Merge Requests that do not have a clear 1:1 mapping - Some MRs have several legacy labels and it's unclear which stage label should be applied - We cannot apply two as stage labels are scoped labels - We can use feature label inference to determine stage labels where possible and manual labelling for the outstanding MRs 3. Determine devops stage label to apply to merged MRs with no legacy label - We can infer the stage labels from the category labels where possible 3. Add devops stage labels to open MRs in the same manner as 1, 2, and 3 - Not essential as only `merged` MRs are counted for throughputs 4. Determine group labels for merged MRs through author ##### Issues 1. Add devops stage labels to issues via feature label inference implemented in https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/155 - We are in the process of rolling this out completely here for both issues and MRs resources - https://gitlab.com/gitlab-org/quality/triage-ops/issues/211

2019-07-04 08:47:24.796

Consider removing the ~group label in favor of the ~groups label

It seems the ~group label (https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=group) is duplicating the ~groups label (https://gitlab.com/groups/gitlab-org/-/issues?scope=all&utf8=✓&state=opened&label_name[]=groups). I suggest removing the ~group label in favor of the ~groups label. Plan of action: - [x] Bulk-add the ~groups label to all the issues/MRs that currently have the ~group label set - [x] Issues: https://gitlab.com/groups/gitlab-org/-/issues?label_name%5B%5D=group - [x] MRs: https://gitlab.com/groups/gitlab-org/-/merge_requests?label_name%5B%5D=group - [x] Epics: https://gitlab.com/groups/gitlab-org/-/epics?label_name%5B%5D=group - [x] Remove the ~group label @jeremy What do you think since you're the product manager for ~"group::access"?

2019-06-24 16:26:45.326

Move all issue rescheduling to happen on the 23rd instead and move it to the next milestone

Captures the following work on rescheduling for https://gitlab.com/gitlab-com/www-gitlab-com/issues/4593 * [x] Feature freeze and missed deliverables https://gitlab.com/gitlab-com/www-gitlab-com/issues/4593#missed-deliverable * Remove feature freeze detection on the 7th * Move missed deliverable when milestone is closed. * Keep bot message * [x] Generic missed items https://gitlab.com/gitlab-com/www-gitlab-com/issues/4593#generic-missed-items-clean-up * Move missed issues and MRs when milestone is closed. * ~~Remove report generation since we are already labeling it with `missed:x.y`~~ * Close the report after generation * Change the report to use numbered list * Fix the bug updating the title for the report * Remove bot message on `missed:x.y` just move the issue

2019-06-20 16:49:29.921

Remove `Deliverable` label if milestone is beyond current+1

The ~Deliverable label is intended to be applied to issues once they have been formally included in a release, gone through kickoff, etc: https://docs.gitlab.com/ee/development/contributing/issue_workflow.html#release-scoping-labels There are issues however where the ~Deliverable is applied to pretty far out issues, for example right now there are 13 ~Deliverable for %"12.4", which has not had any planning rigor. This is proving to frustrate our user community, as this label is now misleading. What has generally happened is the issue was pushed out of a release, and people forgot to remove the label. It would be great if we could remove this automatically, to improve our issue hygiene.

2019-06-18 17:06:50.761

Move generic code to gitlab-triage and keep only domain specific code here

The following discussion from !149 should be addressed: - [ ] @rymai started a [discussion](https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/149#note_182469799): (+2 comments) > 1. `lib/extension.rb` sounds like a very generic name > 1. Why not implementing the generic part of this feature into `gitlab-triage` directly (i.e. `EngineExtension`, `RulePolicyExtension`, `IssueBuilderExtension#initialize`)? That way we'd only override `IssueBuilderExtension#description` to add the heatmap. * [ ] Rename `lib/extension.rb` -> https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/155 * [ ] Introduce another object to contain the conditions and the raw resources -> https://gitlab.com/gitlab-org/gitlab-triage/issues/147 * [ ] Expose the raw resources so we can extend it The most offending code is: ``` ruby yield(resources, conditions.merge(all_resources: all_resources)) ``` @rymai had a comment about it: https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/149#note_182469796 > Could we change this method in `gitlab-triage` to directly yield `resources, conditions, resources_without_limit` instead? I find it hackish to merge `all_resources: all_resources` into `conditions` here... :thinking: To be honest, I spent a lot of time figuring this out, and I got really confused around this code. This was because the object we put into `BasePolicy#resources` can actually have different types for `SummaryPolicy` and `RulePolicy`. I left a comment in `lib/gitlab/triage/policies/summary_policy.rb` ``` ruby # Due to resources is a different type, this will never work # FIXME: We should try to make sure type is consistent for resources def comment? false end ``` This shares the same issue. Depending on the instance, `resources` can mean different things. In `RulePolicy` it's an array of resources, but in `SummaryPolicy` it's a hash mapping from rules to the resources. In order to fix the confusion here, I would like to introduce a new class wrapping it. I don't know how to name it yet, but it should carry the difference. And then, we can attach this `all_resources` (or raw resources) to this object, instead of attaching it to the conditions. I need to attach it to the conditions for now, because I don't know how to properly attach it to either an array or a hash (!) If we try to detect that, it's going to be even more confusing... I consider that a ~"technical debt" without adding an actual object for that part, and I think we need to fix it now before we move on, because we do need to add something on top of it now.

2019-06-18 16:49:45.310

Put issues filter link to the heat map

The following discussion from !149 should be addressed: - [ ] @godfat started a [discussion](https://gitlab.com/gitlab-org/quality/triage-ops/merge_requests/149#note_182700145): > We can definitely construct the issue link here, but I find it not so straightforward because we need to put up the query by ourselves, and this is pretty isolated so I think we can probably do this later.

2023-04-20 17:11:31.408

Add ~"pipeline:run-as-if-jh" when changes are detected in feature flag and dependencies

Follow up from https://gitlab.com/gitlab-org/gitlab/-/merge_requests/109232#note_1361020187 where we cannot describe this in `rules` so we can only do this externally, and via the label is the easiest workaround.

2023-04-13 07:33:43.866

Don't label group/project access token mrs as "community contribution"s

See https://gitlab.com/gitlab-org/secure/vulnerability-research/advisories/cves-private/-/merge_requests/1467 for example I guess we should not be labelling these as community contributions- that way they won't come up in the daily community mr triage report for example /cc @gl-quality/eng-prod

2023-04-06 14:04:50.314

TypeError triage/triage/pipeline_failure/slack_notifier.rb in schedule_type error no implicit conversion of String into Integer

https://sentry.gitlab.net/gitlab/triage-ops/issues/4120780/?environment=production

2023-03-31 19:01:26.909

Enable PipelineFailureManagement for all current use-cases

2023-03-28 14:47:46.637

Introduce SlackNotifier to handle Slack notification for pipeline failures

2023-03-27 10:39:46.945

Introduce classes to create incidents from pipeline events

2023-03-27 10:38:27.695

Port the existing pipeline failure tooling from gitlab-org/gitlab

This is [iteration 1 of the plan](https://gitlab.com/groups/gitlab-org/-/epics/10172#iteration-1).

2023-03-27 10:36:34.256

Support pipeline events

2023-02-09 19:35:11.981

Are we really stubbing everything in devops_labels_spec.rb?

The following discussion from !1949 should be addressed: >>> I have a question though. I thought all of the api calls are stubbed in the spec [here](https://gitlab.com/gitlab-org/quality/triage-ops/-/blob/master/spec/lib/devops_labels_spec.rb#L17), how are we still getting a failing pipeline? I really thought that stubbing would have prevented this type of surprise broken pipelines >>>

2023-01-26 08:48:29.609

gitlab:missed-resources is failing

Job [#3663720487](https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/3663720487) failed for 98d594879ad947c29e81a349e030b755c390282d: ``` W, [2023-01-26T00:26:40.841873 #27] WARN -- sentry: ** [Raven] Failed to submit event: /builds/gitlab-org/quality/triage-ops/lib/missed_resource_helper.rb:42:in `release_date_for': undefined method `year' for nil:NilClass (NoMethodError) from /builds/gitlab-org/quality/triage-ops/lib/missed_resource_helper.rb:9:in `missed_resource?' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/resource/context.rb:31:in `eval' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/resource/context.rb:25:in `instance_eval' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/resource/context.rb:25:in `eval' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/filters/ruby_conditions_filter.rb:22:in `calculate' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:381:in `block in filter_resources' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:333:in `select' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:333:in `filter_resources' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:302:in `block in resources_for_rule' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/expand_condition.rb:13:in `each' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/expand_condition.rb:13:in `perform' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:279:in `resources_for_rule' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:183:in `block in process_rules' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:182:in `each' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:182:in `process_rules' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:79:in `block in perform' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:70:in `each' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/lib/gitlab/triage/engine.rb:70:in `perform' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/bin/gitlab-triage:18:in `block in <top (required)>' from /usr/local/lib/ruby/2.7.0/set.rb:328:in `each_key' from /usr/local/lib/ruby/2.7.0/set.rb:328:in `each' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.28.0/bin/gitlab-triage:11:in `<top (required)>' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/bin/gitlab-triage:23:in `load' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/bin/gitlab-triage:23:in `<main>' ```

2022-08-30 09:24:27.370

Job Failed #2951014112: lib/untriaged_helper.rb:20:in `ceil': NaN (FloatDomainError)

Job [#2951014112](https://gitlab.com/gitlab-org/quality/triage-ops/-/jobs/2951014112) failed for 7a918aa21f56beb7a09da2d530465e07d1e14bb2: Sentry: https://sentry.gitlab.net/gitlab/triage-ops/issues/3430594/?environment=default&referrer=alert_email ``` [DEBUG] rate_limit_infos: Rate limit remaining: 1975 (reset at 2022-08-30 02:08:33 +0000)/builds/gitlab-org/quality/triage-ops/lib/untriaged_helper.rb:20:in `ceil': NaN (FloatDomainError) from /builds/gitlab-org/quality/triage-ops/lib/untriaged_helper.rb:20:in `distribute_items' from /builds/gitlab-org/quality/triage-ops/lib/untriaged_helper.rb:27:in `distribute_and_display_items_per_triager' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/resource/context.rb:82:in `eval' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/resource/context.rb:23:in `instance_eval' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/resource/context.rb:23:in `eval' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/command_builders/text_content_builder.rb:63:in `eval_interpolation' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/command_builders/text_content_builder.rb:53:in `format_item' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/command_builders/base_command_builder.rb:34:in `block in content_string' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/command_builders/base_command_builder.rb:33:in `map' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/command_builders/base_command_builder.rb:33:in `content_string' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/command_builders/base_command_builder.rb:14:in `build_command' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/entity_builders/issue_builder.rb:77:in `build_text' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/entity_builders/issue_builder.rb:30:in `description' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/action/summarize.rb:76:in `post_issue_body' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/action/summarize.rb:44:in `perform' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/action/summarize.rb:33:in `act' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/action.rb:28:in `act' from /builds/gitlab-org/quality/triage-ops/vendor/ruby/2.7.0/gems/gitlab-triage-1.23.1/lib/gitlab/triage/action.rb:16:in `block in process' ```

2022-07-13 13:16:44.204

JiHu AppSec Notifier wasn't triggered on a JiHu Contribution

The [JiHu AppSec notifier](https://about.gitlab.com/handbook/ceo/chief-of-staff-team/jihu-support/jihu-security-review-process.html) was not triggered on https://gitlab.com/gitlab-org/gitlab/-/merge_requests/90956#note_a8f30383f81d5e765f9e0132e3f5d6d4d32bcd80 and we need to find out why.

2022-07-12 14:36:05.139

Remind MR authors that set MWPS that this is not a compliant action

## Objective Reduce the amount of MRs merged to `gitlab-org/gitlab` by the MR author as apart of https://gitlab.com/gitlab-com/gl-security/security-assurance/security-compliance-commercial-and-dedicated/sec-compliance/observation-management/-/issues/318 ## Proposal Remind the MR Author (similar to https://gitlab.com/gitlab-org/gitlab/-/merge_requests/89719#note_994821846) when they set MWPS on their MR that this is a non-compliant action and another maintainer should merge the MR. Use https://gitlab.com/gitlab-org/quality/triage-ops/-/blob/master/triage/processor/remind_merged_mr_deviating_from_guidelines.rb as a reference implementation on the merge when pipeline succeeds event. ### Limitations - Approvals can't be revoked (https://gitlab.com/gitlab-org/gitlab/-/issues/367392) so we are going to inform first and then block merging with https://gitlab.com/gitlab-org/gitlab/-/issues/366696

2022-05-24 13:30:20.277

Don't ask community contributions to set a type label for now

It seems we're missing the logic to exclude ~"Community contribution" in https://gitlab.com/gitlab-org/quality/triage-ops/-/blob/d7baaeed5b5b32ad8e8046cb196ec0e5838ee605/triage/job/type_label_nudger_job.rb#L17. See https://gitlab.com/gitlab-org/gitlab/-/merge_requests/88409#note_958051092 for an example.

2022-05-20 12:05:53.659

Suggest community contributors to use gitlab-org/gitlab on creating an MR on gitlab-org/gitlab-foss

`@gitlab-bot` should suggest community contributors to use gitlab-org/gitlab on creating an MR on `gitlab-org/gitlab-foss`. Currently bot suggests the same message with one for `gitlab-org/gitlab`. ### Example https://gitlab.com/gitlab-org/gitlab-foss/-/merge_requests/33104#note_954012005

2022-05-19 10:35:30.297

Consider making the `request_review` reactive command response message more concise

The response message for the `request_review` reactive command is currently a bit verbose I think. Example of the current behavior (https://gitlab.com/gitlab-org/gitlab/-/merge_requests/87857#note_951003725):  Making it more concise would reduce the noise for ~"Community contribution" authors (and everyone else).

2022-05-11 08:00:40.484

Group community-related processors under `triage/processor/community`

2022-05-06 10:37:45.762

Auto apply Hackathon label and link with tracking issue for FY23Q2 Hackathon

## Summary This issue is for requesting triage automation on a given project. Please link to the project needed and select from a list of triage automation rules to be added. Project: https://gitlab.com/gitlab-com, https://gitlab.com/gitlab-org * Automation that does not require an assignee. * Labelling ~"Hackathon" on any ~"Community contribution" labeled MR during May 9th, 12 pm UTC (noon), and May 13th 12 pm UTC (noon) * Leaving the following comment: ` This MR will be considered [part of](https://gitlab.com/gitlab-com/marketing/community-relations/contributor-program/hackathon/-/issues/65) the quarterly [GitLab Hackathon](https://about.gitlab.com/community/hackathon/) for a chance to win a [prize](https://about.gitlab.com/community/hackathon/#prize). Can you make sure this MR is mentioning/linking the relevant it's attempting to close? Thank you for your contribution!` cc @kwiebers

2022-05-03 09:55:11.452

Implement type label nudger processor

## Objective Reduce maintenance for updates to type label nudge and allow for easier rollout to projects ### Why is this important? The 3 types (Bug, Feature & Maintenance) is key to our report to industry analysts. It is important for GitLab to communicate effort spent into a format that is easily understandable widely in the industry. We provide this metric to our leadership reporting and improve the accuracy with subtypes categorization. Additionally tracking accurate work output allows the team to compare output to backlog and adjust focus accordingly. ### Proposal 1. Create a type label discussion for non-\~"Community contribution" MRs opened without a type for the [top 20 offending undefined projects](https://app.periscopedata.com/app/gitlab/1021219/Type-Research?widget=14627996&udv=0) with the below message requesting the MR be classified label. => https://gitlab.com/gitlab-org/quality/triage-ops/-/merge_requests/1374 1. Expand to additional projects based on [undefined ratio](https://app.periscopedata.com/app/gitlab/1021219/Type-Research?widget=14627996&udv=0) (what proportion of merged MRs do not have a type) for projects over 20 MRs in the last 90 days 1. Expand to all `is_part_of_product` projects ``` :wave: @#{event.event_actor_username} - please add ~"type::bug", ~"type::feature", ~"type::maintenance", or a [subtype](https://about.gitlab.com/handbook/engineering/metrics/#work-type-classification) label to this merge request. - ~"type::bug": Defects in shipped code and fixes for those defects. This includes all the bug types (availability, performance, security vulnerability, mobile, etc.) - ~"type::feature": Effort to deliver new features, feature changes & improvements. This includes all changes as part of new product requirements like application limits. - ~"type::maintenance"`: Up-keeping efforts & catch-up corrective improvements that are not Features nor Bugs. This includes restructuring for long-term maintainability, stability, reducing technical debt, improving the contributor experience, or upgrading dependencies. See [the handbook](https://about.gitlab.com/handbook/engineering/metrics/#work-type-classification) for more guidance on classifying. ```

2022-04-25 17:30:02.127

Replace SUS with non-sus label when feature::addition or Actionable Insight::exploration needed is applied to issues

## Objective Replace ~SUS with ~"non-SUS" if ~"feature::addition" or ~"Actionable Insight::Exploration needed" is on an issue so that the rules in https://gitlab.com/gitlab-com/www-gitlab-com/-/merge_requests/100632#note_918555284 can be automated. ## Acceptance Criteria * Convert ~SUS labels to be scoped: https://gitlab.com/gitlab-org/quality/triage-ops/-/issues/982#note_923537080

2022-04-22 11:09:44.718

Gitlab::Error::BadRequest: Server responded with code 400, message: 400 Bad request - Note {:note=>["can't be blank"]}. Requ...

Sentry Issue: [TRIAGE-OPS-DA](https://sentry.gitlab.net/gitlab/triage-ops/issues/3270725/?referrer=gitlab_integration) ``` Gitlab::Error::BadRequest: Server responded with code 400, message: 400 Bad request - Note {:note=>["can't be blank"]}. Request URI: https://gitlab.com/api/v4/projects/20699/merge_requests/6055/notes gitlab/request.rb:71:in `validate' raise error_klass, response if error_klass gitlab/request.rb:55:in `block (2 levels) in <class:Request>' validate response triage/triage/reaction.rb:40:in `post_request' Triage.api_client.post(path, body: { body: body }) unless Triage.dry_run? triage/triage/reaction.rb:12:in `add_comment' Reaction.post_request(path, body) triage/processor/reactive_labeler.rb:50:in `post_label_command' add_comment <<~MARKDOWN.chomp ... (16 additional frame(s) were not displayed) ``` I believe the issue happens when someone asks to apply a label that doesn't match the allowed pattern, e.g. `@gitlab-bot label \~"group::Omnibus Package"`.