forgejo/modules/indexer/issues/bleve.go
wxiaoguang 6f9c278559
Rewrite queue (#24505)
# ⚠️ Breaking

Many deprecated queue config options are removed (actually, they should
have been removed in 1.18/1.19).

If you see the fatal message when starting Gitea: "Please update your
app.ini to remove deprecated config options", please follow the error
messages to remove these options from your app.ini.

Example:

```
2023/05/06 19:39:22 [E] Removed queue option: `[indexer].ISSUE_INDEXER_QUEUE_TYPE`. Use new options in `[queue.issue_indexer]`
2023/05/06 19:39:22 [E] Removed queue option: `[indexer].UPDATE_BUFFER_LEN`. Use new options in `[queue.issue_indexer]`
2023/05/06 19:39:22 [F] Please update your app.ini to remove deprecated config options
```

Many options in `[queue]` are are dropped, including:
`WRAP_IF_NECESSARY`, `MAX_ATTEMPTS`, `TIMEOUT`, `WORKERS`,
`BLOCK_TIMEOUT`, `BOOST_TIMEOUT`, `BOOST_WORKERS`, they can be removed
from app.ini.

# The problem

The old queue package has some legacy problems:

* complexity: I doubt few people could tell how it works.
* maintainability: Too many channels and mutex/cond are mixed together,
too many different structs/interfaces depends each other.
* stability: due to the complexity & maintainability, sometimes there
are strange bugs and difficult to debug, and some code doesn't have test
(indeed some code is difficult to test because a lot of things are mixed
together).
* general applicability: although it is called "queue", its behavior is
not a well-known queue.
* scalability: it doesn't seem easy to make it work with a cluster
without breaking its behaviors.

It came from some very old code to "avoid breaking", however, its
technical debt is too heavy now. It's a good time to introduce a better
"queue" package.

# The new queue package

It keeps using old config and concept as much as possible.

* It only contains two major kinds of concepts:
    * The "base queue": channel, levelqueue, redis
* They have the same abstraction, the same interface, and they are
tested by the same testing code.
* The "WokerPoolQueue", it uses the "base queue" to provide "worker
pool" function, calls the "handler" to process the data in the base
queue.
* The new code doesn't do "PushBack"
* Think about a queue with many workers, the "PushBack" can't guarantee
the order for re-queued unhandled items, so in new code it just does
"normal push"
* The new code doesn't do "pause/resume"
* The "pause/resume" was designed to handle some handler's failure: eg:
document indexer (elasticsearch) is down
* If a queue is paused for long time, either the producers blocks or the
new items are dropped.
* The new code doesn't do such "pause/resume" trick, it's not a common
queue's behavior and it doesn't help much.
* If there are unhandled items, the "push" function just blocks for a
few seconds and then re-queue them and retry.
* The new code doesn't do "worker booster"
* Gitea's queue's handlers are light functions, the cost is only the
go-routine, so it doesn't make sense to "boost" them.
* The new code only use "max worker number" to limit the concurrent
workers.
* The new "Push" never blocks forever
* Instead of creating more and more blocking goroutines, return an error
is more friendly to the server and to the end user.

There are more details in code comments: eg: the "Flush" problem, the
strange "code.index" hanging problem, the "immediate" queue problem.

Almost ready for review.

TODO:

* [x] add some necessary comments during review
* [x] add some more tests if necessary
* [x] update documents and config options
* [x] test max worker / active worker
* [x] re-run the CI tasks to see whether any test is flaky
* [x] improve the `handleOldLengthConfiguration` to provide more
friendly messages
* [x] fine tune default config values (eg: length?)

## Code coverage:

![image](https://user-images.githubusercontent.com/2114189/236620635-55576955-f95d-4810-b12f-879026a3afdf.png)
2023-05-08 19:49:59 +08:00

276 lines
7.6 KiB
Go

// Copyright 2018 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package issues
import (
"context"
"fmt"
"os"
"strconv"
gitea_bleve "code.gitea.io/gitea/modules/indexer/bleve"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/util"
"github.com/blevesearch/bleve/v2"
"github.com/blevesearch/bleve/v2/analysis/analyzer/custom"
"github.com/blevesearch/bleve/v2/analysis/token/camelcase"
"github.com/blevesearch/bleve/v2/analysis/token/lowercase"
"github.com/blevesearch/bleve/v2/analysis/token/unicodenorm"
"github.com/blevesearch/bleve/v2/analysis/tokenizer/unicode"
"github.com/blevesearch/bleve/v2/index/upsidedown"
"github.com/blevesearch/bleve/v2/mapping"
"github.com/blevesearch/bleve/v2/search/query"
"github.com/ethantkoenig/rupture"
)
const (
issueIndexerAnalyzer = "issueIndexer"
issueIndexerDocType = "issueIndexerDocType"
issueIndexerLatestVersion = 2
)
// indexerID a bleve-compatible unique identifier for an integer id
func indexerID(id int64) string {
return strconv.FormatInt(id, 36)
}
// idOfIndexerID the integer id associated with an indexer id
func idOfIndexerID(indexerID string) (int64, error) {
id, err := strconv.ParseInt(indexerID, 36, 64)
if err != nil {
return 0, fmt.Errorf("Unexpected indexer ID %s: %w", indexerID, err)
}
return id, nil
}
// numericEqualityQuery a numeric equality query for the given value and field
func numericEqualityQuery(value int64, field string) *query.NumericRangeQuery {
f := float64(value)
tru := true
q := bleve.NewNumericRangeInclusiveQuery(&f, &f, &tru, &tru)
q.SetField(field)
return q
}
func newMatchPhraseQuery(matchPhrase, field, analyzer string) *query.MatchPhraseQuery {
q := bleve.NewMatchPhraseQuery(matchPhrase)
q.FieldVal = field
q.Analyzer = analyzer
return q
}
const unicodeNormalizeName = "unicodeNormalize"
func addUnicodeNormalizeTokenFilter(m *mapping.IndexMappingImpl) error {
return m.AddCustomTokenFilter(unicodeNormalizeName, map[string]interface{}{
"type": unicodenorm.Name,
"form": unicodenorm.NFC,
})
}
const maxBatchSize = 16
// openIndexer open the index at the specified path, checking for metadata
// updates and bleve version updates. If index needs to be created (or
// re-created), returns (nil, nil)
func openIndexer(path string, latestVersion int) (bleve.Index, error) {
_, err := os.Stat(path)
if err != nil && os.IsNotExist(err) {
return nil, nil
} else if err != nil {
return nil, err
}
metadata, err := rupture.ReadIndexMetadata(path)
if err != nil {
return nil, err
}
if metadata.Version < latestVersion {
// the indexer is using a previous version, so we should delete it and
// re-populate
return nil, util.RemoveAll(path)
}
index, err := bleve.Open(path)
if err != nil && err == upsidedown.IncompatibleVersion {
// the indexer was built with a previous version of bleve, so we should
// delete it and re-populate
return nil, util.RemoveAll(path)
} else if err != nil {
return nil, err
}
return index, nil
}
// BleveIndexerData an update to the issue indexer
type BleveIndexerData IndexerData
// Type returns the document type, for bleve's mapping.Classifier interface.
func (i *BleveIndexerData) Type() string {
return issueIndexerDocType
}
// createIssueIndexer create an issue indexer if one does not already exist
func createIssueIndexer(path string, latestVersion int) (bleve.Index, error) {
mapping := bleve.NewIndexMapping()
docMapping := bleve.NewDocumentMapping()
numericFieldMapping := bleve.NewNumericFieldMapping()
numericFieldMapping.IncludeInAll = false
docMapping.AddFieldMappingsAt("RepoID", numericFieldMapping)
textFieldMapping := bleve.NewTextFieldMapping()
textFieldMapping.Store = false
textFieldMapping.IncludeInAll = false
docMapping.AddFieldMappingsAt("Title", textFieldMapping)
docMapping.AddFieldMappingsAt("Content", textFieldMapping)
docMapping.AddFieldMappingsAt("Comments", textFieldMapping)
if err := addUnicodeNormalizeTokenFilter(mapping); err != nil {
return nil, err
} else if err = mapping.AddCustomAnalyzer(issueIndexerAnalyzer, map[string]interface{}{
"type": custom.Name,
"char_filters": []string{},
"tokenizer": unicode.Name,
"token_filters": []string{unicodeNormalizeName, camelcase.Name, lowercase.Name},
}); err != nil {
return nil, err
}
mapping.DefaultAnalyzer = issueIndexerAnalyzer
mapping.AddDocumentMapping(issueIndexerDocType, docMapping)
mapping.AddDocumentMapping("_all", bleve.NewDocumentDisabledMapping())
index, err := bleve.New(path, mapping)
if err != nil {
return nil, err
}
if err = rupture.WriteIndexMetadata(path, &rupture.IndexMetadata{
Version: latestVersion,
}); err != nil {
return nil, err
}
return index, nil
}
var _ Indexer = &BleveIndexer{}
// BleveIndexer implements Indexer interface
type BleveIndexer struct {
indexDir string
indexer bleve.Index
}
// NewBleveIndexer creates a new bleve local indexer
func NewBleveIndexer(indexDir string) *BleveIndexer {
return &BleveIndexer{
indexDir: indexDir,
}
}
// Init will initialize the indexer
func (b *BleveIndexer) Init() (bool, error) {
var err error
b.indexer, err = openIndexer(b.indexDir, issueIndexerLatestVersion)
if err != nil {
return false, err
}
if b.indexer != nil {
return true, nil
}
b.indexer, err = createIssueIndexer(b.indexDir, issueIndexerLatestVersion)
return false, err
}
// Ping does nothing
func (b *BleveIndexer) Ping() bool {
return true
}
// Close will close the bleve indexer
func (b *BleveIndexer) Close() {
if b.indexer != nil {
if err := b.indexer.Close(); err != nil {
log.Error("Error whilst closing indexer: %v", err)
}
}
}
// Index will save the index data
func (b *BleveIndexer) Index(issues []*IndexerData) error {
batch := gitea_bleve.NewFlushingBatch(b.indexer, maxBatchSize)
for _, issue := range issues {
if err := batch.Index(indexerID(issue.ID), struct {
RepoID int64
Title string
Content string
Comments []string
}{
RepoID: issue.RepoID,
Title: issue.Title,
Content: issue.Content,
Comments: issue.Comments,
}); err != nil {
return err
}
}
return batch.Flush()
}
// Delete deletes indexes by ids
func (b *BleveIndexer) Delete(ids ...int64) error {
batch := gitea_bleve.NewFlushingBatch(b.indexer, maxBatchSize)
for _, id := range ids {
if err := batch.Delete(indexerID(id)); err != nil {
return err
}
}
return batch.Flush()
}
// Search searches for issues by given conditions.
// Returns the matching issue IDs
func (b *BleveIndexer) Search(ctx context.Context, keyword string, repoIDs []int64, limit, start int) (*SearchResult, error) {
var repoQueriesP []*query.NumericRangeQuery
for _, repoID := range repoIDs {
repoQueriesP = append(repoQueriesP, numericEqualityQuery(repoID, "RepoID"))
}
repoQueries := make([]query.Query, len(repoQueriesP))
for i, v := range repoQueriesP {
repoQueries[i] = query.Query(v)
}
indexerQuery := bleve.NewConjunctionQuery(
bleve.NewDisjunctionQuery(repoQueries...),
bleve.NewDisjunctionQuery(
newMatchPhraseQuery(keyword, "Title", issueIndexerAnalyzer),
newMatchPhraseQuery(keyword, "Content", issueIndexerAnalyzer),
newMatchPhraseQuery(keyword, "Comments", issueIndexerAnalyzer),
))
search := bleve.NewSearchRequestOptions(indexerQuery, limit, start, false)
search.SortBy([]string{"-_score"})
result, err := b.indexer.SearchInContext(ctx, search)
if err != nil {
return nil, err
}
ret := SearchResult{
Hits: make([]Match, 0, len(result.Hits)),
}
for _, hit := range result.Hits {
id, err := idOfIndexerID(hit.ID)
if err != nil {
return nil, err
}
ret.Hits = append(ret.Hits, Match{
ID: id,
})
}
return &ret, nil
}