fix: realign indexes on the 'action' table (#10040)

Fixes #9963.  This realigns all the indexes on the `action` table to best match their intended usages.

New:
- `IDX_action_created_unix (created_unix)`
    - Intended for usage in `DeleteOldActions`.
- `IDX_action_repo_id_created_unix (repo_id, created_unix)`
    - Intended for usage when fetching action feeds for a repo and a team, with the same logic as that described below in `IDX_action_user_id_created_unix`.
- `IDX_action_repo_id_op_type (repo_id, op_type)`
    - Intended for `DeleteIssueActions` when it searches for CreateIssue & CreatePullRequest actions for cleanup.  Could be optimized further with a denormalization of the issue identifier into a field, but there's no current evidence that this is required.

Replaced:
- `IDX_action_c_u (created_unix, user_id)`
    - Replaced with `IDX_action_user_id_created_unix (user_id, created_unix)`.  When action feeds are created w/ `ORDER BY created_unix DESC LIMIT 20`, an index beginning with `created_unix` will have to index scan until it can satisfy 20 records; the `user_id` portion of the index is effectively useless until two records appear at the same time.  By inverting the order, the database will be able to identify the records created by a user and then pop the most recent 20 from the index order.
    - At the scale of database I have access to, the performance difference is unmeasurable.  This change is supported by theoretical grounds and the findings of #9963, but no experimental evidence.

Removed:
- `IDX_action_user_id (user_id)`
    - Redundant with the new `IDX_action_user_id_created_unix`.
- `IDX_action_r_u (repo_id, user_id)`
    - No clear consumer for this index.

Retained with no modification:
- `IDX_action_comment_id (comment_id)`
    - Used in `DeleteIssueActions`.
- `IDX_action_au_r_c_u (act_user_id, repo_id, created_unix, user_id)`
    - Heat map generation.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10040
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
This commit is contained in:
Mathieu Fenniak 2025-11-18 18:34:25 +01:00 committed by Mathieu Fenniak
parent afbd05c398
commit dea9ef6706
4 changed files with 209 additions and 10 deletions

View file

@ -145,37 +145,49 @@ func (at ActionType) InActions(actions ...string) bool {
// used in template render.
type Action struct {
ID int64 `xorm:"pk autoincr"`
UserID int64 `xorm:"INDEX"` // Receiver user id.
UserID int64 // Receiver user id.
OpType ActionType
ActUserID int64 // Action user id.
ActUser *user_model.User `xorm:"-"`
RepoID int64
Repo *repo_model.Repository `xorm:"-"`
CommentID int64 `xorm:"INDEX"`
CommentID int64 `xorm:"INDEX"` // indexed to support `DeleteIssueActions`
Comment *issues_model.Comment `xorm:"-"`
Issue *issues_model.Issue `xorm:"-"` // get the issue id from content
RefName string
IsPrivate bool `xorm:"NOT NULL DEFAULT false"`
Content string `xorm:"TEXT"`
CreatedUnix timeutil.TimeStamp `xorm:"created"`
CreatedUnix timeutil.TimeStamp `xorm:"created INDEX"` // indexed to support `DeleteOldActions`
}
func init() {
db.RegisterModel(new(Action))
}
// TableIndices implements xorm's TableIndices interface
// TableIndices implements xorm's TableIndices interface. It is used here to ensure indexes with specified column order
// are created, which can't be created through xorm tags on the struct.
func (a *Action) TableIndices() []*schemas.Index {
repoIndex := schemas.NewIndex("r_u", schemas.IndexType)
repoIndex.AddColumn("repo_id", "user_id")
// Index to support getUserHeatmapData, which searches for data that is visible-to (user_id) and performed-by
// (act_user_id) a user, but only includes visible repos (repo_id).
actUserIndex := schemas.NewIndex("au_r_c_u", schemas.IndexType)
actUserIndex.AddColumn("act_user_id", "repo_id", "created_unix", "user_id")
cudIndex := schemas.NewIndex("c_u", schemas.IndexType)
cudIndex.AddColumn("created_unix", "user_id")
// GetFeeds is a common access point to Action and requires that all action feeds be queried based upon one of
// user_id (opts.RequestedUser), repo_id (opts.RequestedTeam... kinda), and/or repo_id (opts.RequestedRepo), and
// then the results are ordered by created_unix and paginated. The most efficient indexes to support those queries
// are:
requestedUser := schemas.NewIndex("user_id_created_unix", schemas.IndexType)
requestedUser.AddColumn("user_id", "created_unix")
requestedRepo := schemas.NewIndex("repo_id_created_unix", schemas.IndexType)
requestedRepo.AddColumn("repo_id", "created_unix")
indices := []*schemas.Index{actUserIndex, repoIndex, cudIndex}
// To support `DeleteIssueActions` search for createissue / createpullrequest actions; this isn't a great search
// because `DeleteIssueActions` searches by `content` as well, but it should be sufficient performance-wise for
// infrequent deleting of issues.
repoOpType := schemas.NewIndex("repo_id_op_type", schemas.IndexType)
repoOpType.AddColumn("repo_id", "op_type")
indices := []*schemas.Index{actUserIndex, requestedUser, requestedRepo, repoOpType}
return indices
}