This commit is contained in:
Max Ignatenko 2024-02-15 16:10:39 +00:00
parent 2b6abac607
commit 63a767d890
25 changed files with 3027 additions and 0 deletions

14
cmd/lister/Dockerfile Normal file
View file

@ -0,0 +1,14 @@
FROM golang:1.21.1 as builder
WORKDIR /app
COPY go.mod go.sum ./
RUN go mod download
COPY . ./
RUN go build -trimpath ./cmd/lister
FROM alpine:latest as certs
RUN apk --update add ca-certificates
FROM debian:stable-slim
COPY --from=certs /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/ca-certificates.crt
COPY --from=builder /app/lister .
ENTRYPOINT ["./lister"]

108
cmd/lister/lister.go Normal file
View file

@ -0,0 +1,108 @@
package main
import (
"context"
"errors"
"time"
"github.com/rs/zerolog"
"gorm.io/gorm"
comatproto "github.com/bluesky-social/indigo/api/atproto"
"github.com/bluesky-social/indigo/did"
"github.com/uabluerail/bsky-tools/pagination"
"github.com/uabluerail/bsky-tools/xrpcauth"
"github.com/uabluerail/indexer/pds"
"github.com/uabluerail/indexer/repo"
"github.com/uabluerail/indexer/util/resolver"
)
type Lister struct {
db *gorm.DB
resolver did.Resolver
pollInterval time.Duration
listRefreshInterval time.Duration
}
func NewLister(ctx context.Context, db *gorm.DB) (*Lister, error) {
return &Lister{
db: db,
resolver: resolver.Resolver,
pollInterval: 5 * time.Minute,
listRefreshInterval: 24 * time.Hour,
}, nil
}
func (l *Lister) Start(ctx context.Context) error {
go l.run(ctx)
return nil
}
func (l *Lister) run(ctx context.Context) {
log := zerolog.Ctx(ctx)
ticker := time.NewTicker(l.pollInterval)
log.Info().Msgf("Lister starting...")
t := make(chan time.Time, 1)
t <- time.Now()
for {
select {
case <-ctx.Done():
log.Info().Msgf("Lister stopped (context expired)")
return
case <-t:
db := l.db.WithContext(ctx)
remote := pds.PDS{}
if err := db.Model(&remote).
Where("last_list is null or last_list < ?", time.Now().Add(-l.listRefreshInterval)).
Take(&remote).Error; err != nil {
if !errors.Is(err, gorm.ErrRecordNotFound) {
log.Error().Err(err).Msgf("Failed to query DB for a PDS to list repos from: %s", err)
}
break
}
client := xrpcauth.NewAnonymousClient(ctx)
client.Host = remote.Host
log.Info().Msgf("Listing repos from %q...", remote.Host)
dids, err := pagination.Reduce(
func(cursor string) (resp *comatproto.SyncListRepos_Output, nextCursor string, err error) {
resp, err = comatproto.SyncListRepos(ctx, client, cursor, 200)
if err == nil && resp.Cursor != nil {
nextCursor = *resp.Cursor
}
return
},
func(resp *comatproto.SyncListRepos_Output, acc []string) ([]string, error) {
for _, repo := range resp.Repos {
if repo == nil {
continue
}
acc = append(acc, repo.Did)
}
return acc, nil
})
if err != nil {
log.Error().Err(err).Msgf("Failed to list repos from %q: %s", remote.Host, err)
break
}
log.Info().Msgf("Received %d DIDs from %q", len(dids), remote.Host)
for _, did := range dids {
if _, err := repo.EnsureExists(ctx, l.db, did); err != nil {
log.Error().Err(err).Msgf("Failed to ensure that we have a record for the repo %q: %s", did, err)
}
}
if err := db.Model(&remote).Updates(&pds.PDS{LastList: time.Now()}).Error; err != nil {
log.Error().Err(err).Msgf("Failed to update the timestamp of last list for %q: %s", remote.Host, err)
}
case v := <-ticker.C:
t <- v
}
}
}

180
cmd/lister/main.go Normal file
View file

@ -0,0 +1,180 @@
package main
import (
"context"
"flag"
"fmt"
"io"
"log"
"net/http"
_ "net/http/pprof"
"os"
"os/signal"
"path/filepath"
"runtime"
"runtime/debug"
"strings"
"syscall"
"time"
_ "github.com/joho/godotenv/autoload"
"github.com/kelseyhightower/envconfig"
"github.com/prometheus/client_golang/prometheus/promhttp"
"github.com/rs/zerolog"
"gorm.io/driver/postgres"
"gorm.io/gorm"
"gorm.io/gorm/logger"
"github.com/uabluerail/indexer/pds"
"github.com/uabluerail/indexer/repo"
"github.com/uabluerail/indexer/util/gormzerolog"
)
type Config struct {
LogFile string
LogFormat string `default:"text"`
LogLevel int64 `default:"1"`
MetricsPort string `split_words:"true"`
DBUrl string `envconfig:"POSTGRES_URL"`
}
var config Config
func runMain(ctx context.Context) error {
ctx = setupLogging(ctx)
log := zerolog.Ctx(ctx)
log.Debug().Msgf("Starting up...")
db, err := gorm.Open(postgres.Open(config.DBUrl), &gorm.Config{
Logger: gormzerolog.New(&logger.Config{
SlowThreshold: 1 * time.Second,
IgnoreRecordNotFoundError: true,
}, nil),
})
if err != nil {
return fmt.Errorf("connecting to the database: %w", err)
}
log.Debug().Msgf("DB connection established")
for _, f := range []func(*gorm.DB) error{
pds.AutoMigrate,
repo.AutoMigrate,
} {
if err := f(db); err != nil {
return fmt.Errorf("auto-migrating DB schema: %w", err)
}
}
log.Debug().Msgf("DB schema updated")
lister, err := NewLister(ctx, db)
if err != nil {
return fmt.Errorf("failed to create lister: %w", err)
}
if err := lister.Start(ctx); err != nil {
return fmt.Errorf("failed to start lister: %w", err)
}
log.Info().Msgf("Starting HTTP listener on %q...", config.MetricsPort)
http.Handle("/metrics", promhttp.Handler())
srv := &http.Server{Addr: fmt.Sprintf(":%s", config.MetricsPort)}
errCh := make(chan error)
go func() {
errCh <- srv.ListenAndServe()
}()
select {
case <-ctx.Done():
if err := srv.Shutdown(context.Background()); err != nil {
return fmt.Errorf("HTTP server shutdown failed: %w", err)
}
}
return <-errCh
}
func main() {
flag.StringVar(&config.LogFile, "log", "", "Path to the log file. If empty, will log to stderr")
flag.StringVar(&config.LogFormat, "log-format", "text", "Logging format. 'text' or 'json'")
flag.Int64Var(&config.LogLevel, "log-level", 1, "Log level. -1 - trace, 0 - debug, 1 - info, 5 - panic")
if err := envconfig.Process("lister", &config); err != nil {
log.Fatalf("envconfig.Process: %s", err)
}
flag.Parse()
ctx, _ := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM)
if err := runMain(ctx); err != nil {
fmt.Fprintln(os.Stderr, err)
os.Exit(1)
}
}
func setupLogging(ctx context.Context) context.Context {
logFile := os.Stderr
if config.LogFile != "" {
f, err := os.OpenFile(config.LogFile, os.O_APPEND|os.O_CREATE|os.O_WRONLY, 0644)
if err != nil {
log.Fatalf("Failed to open the specified log file %q: %s", config.LogFile, err)
}
logFile = f
}
var output io.Writer
switch config.LogFormat {
case "json":
output = logFile
case "text":
prefixList := []string{}
info, ok := debug.ReadBuildInfo()
if ok {
prefixList = append(prefixList, info.Path+"/")
}
basedir := ""
_, sourceFile, _, ok := runtime.Caller(0)
if ok {
basedir = filepath.Dir(sourceFile)
}
if basedir != "" && strings.HasPrefix(basedir, "/") {
prefixList = append(prefixList, basedir+"/")
head, _ := filepath.Split(basedir)
for head != "/" {
prefixList = append(prefixList, head)
head, _ = filepath.Split(strings.TrimSuffix(head, "/"))
}
}
output = zerolog.ConsoleWriter{
Out: logFile,
NoColor: true,
TimeFormat: time.RFC3339,
PartsOrder: []string{
zerolog.LevelFieldName,
zerolog.TimestampFieldName,
zerolog.CallerFieldName,
zerolog.MessageFieldName,
},
FormatFieldName: func(i interface{}) string { return fmt.Sprintf("%s:", i) },
FormatFieldValue: func(i interface{}) string { return fmt.Sprintf("%s", i) },
FormatCaller: func(i interface{}) string {
s := i.(string)
for _, p := range prefixList {
s = strings.TrimPrefix(s, p)
}
return s
},
}
default:
log.Fatalf("Invalid log format specified: %q", config.LogFormat)
}
logger := zerolog.New(output).Level(zerolog.Level(config.LogLevel)).With().Caller().Timestamp().Logger()
ctx = logger.WithContext(ctx)
zerolog.DefaultContextLogger = &logger
log.SetOutput(logger)
return ctx
}