Files
restic-manager/internal/agent/runner/runner.go
T
steve 3904a78f14
CI / Test (linux/amd64) (push) Has been cancelled
CI / Lint (push) Has been cancelled
CI / Build (windows/amd64) (push) Has been cancelled
CI / Build (linux/amd64) (push) Has been cancelled
CI / Build (linux/arm64) (push) Has been cancelled
P1-22: snapshot listing via restic snapshots --json
Agent calls restic snapshots --json after each successful backup
(60s timeout, separate from the backup ctx) and ships the projection
over the existing snapshots.report WS envelope. Failure here is
logged but doesn't fail the job — the next successful backup catches
the projection up.

Server-side ReplaceHostSnapshots is delete-then-insert plus a
hosts.snapshot_count update in one transaction so the dashboard's
per-host count stays consistent with the projection. New read
endpoint GET /api/hosts/{id}/snapshots returns the cached list with
a refreshed_at marker so the UI can show staleness when an agent
has been offline.

Schema: dropped the unused snapshots.repo_id FK (repos as a
first-class entity is P2 work), added short_id and refreshed_at
columns, switched the time index to DESC for the most-recent-first
list query. api.Snapshot gains short_id; size_bytes/file_count come
from the embedded summary block on restic 0.16+ and stay zero on
older clients.

Tests cover round-trip, authoritative replacement after forget+prune
shrinkage, and empty-after-wipe.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-01 11:20:57 +01:00

190 lines
5.3 KiB
Go

// Package runner spawns restic processes for the agent. It owns one
// Run() invocation per command.run; concurrency limits live a layer
// up (the WS handler).
package runner
import (
"context"
"encoding/json"
"fmt"
"log/slog"
"sync/atomic"
"time"
"gitea.dcglab.co.uk/steve/restic-manager/internal/api"
"gitea.dcglab.co.uk/steve/restic-manager/internal/restic"
)
// Sender is the agent's outbound message channel. Provided by
// wsclient so the runner can push job.started / job.progress /
// job.finished / log.stream back to the server.
type Sender interface {
Send(env api.Envelope) error
}
// Config bundles the long-lived settings the runner needs. They come
// from the agent's config file (server-pushed config.update payloads
// override these in memory).
type Config struct {
ResticBin string
RepoURL string
RepoPassword string
}
// Runner owns the restic invocations.
type Runner struct {
cfg Config
tx Sender
// progress throttling: we receive a status event from restic
// every ~100ms; the UI doesn't need anywhere near that rate.
// Cap WS sends to one per N (configurable; default 1s).
progressMinPeriod time.Duration
}
// New builds a Runner. progressMinPeriod = 0 uses the default 1s.
func New(cfg Config, tx Sender, progressMinPeriod time.Duration) *Runner {
if progressMinPeriod <= 0 {
progressMinPeriod = time.Second
}
return &Runner{cfg: cfg, tx: tx, progressMinPeriod: progressMinPeriod}
}
// RunBackup executes a backup job and reports back via the sender.
// Returns nil on a clean (or "incomplete-but-snapshot-created") finish.
func (r *Runner) RunBackup(ctx context.Context, jobID string, paths, excludes, tags []string) error {
startedAt := time.Now().UTC()
startEnv, _ := api.Marshal(api.MsgJobStarted, jobID, api.JobStartedPayload{
JobID: jobID, Kind: api.JobBackup, StartedAt: startedAt,
})
if err := r.tx.Send(startEnv); err != nil {
slog.Warn("runner: send job.started", "err", err)
}
env := restic.Env{
Bin: r.cfg.ResticBin,
RepoURL: r.cfg.RepoURL,
RepoPassword: r.cfg.RepoPassword,
}
var seq atomic.Int64
lastProgress := time.Now()
handle := func(stream string, line string, ev any) {
// Forward every line to the server as log.stream.
now := time.Now().UTC()
logEnv, _ := api.Marshal(api.MsgLogStream, "", api.LogStreamLine{
JobID: jobID,
Seq: seq.Add(1),
TS: now,
Stream: api.LogStream(stream),
Payload: line,
})
_ = r.tx.Send(logEnv)
// Throttled progress events.
if status, ok := ev.(restic.BackupStatus); ok {
if time.Since(lastProgress) < r.progressMinPeriod {
return
}
lastProgress = time.Now()
progEnv, _ := api.Marshal(api.MsgJobProgress, jobID, api.JobProgressPayload{
JobID: jobID,
PercentDone: status.PercentDone,
FilesDone: status.FilesDone,
TotalFiles: status.TotalFiles,
BytesDone: status.BytesDone,
TotalBytes: status.TotalBytes,
ETASeconds: status.SecondsRem,
ThroughputBps: throughput(status.BytesDone, status.SecondsElapsed),
})
_ = r.tx.Send(progEnv)
}
}
summary, err := env.RunBackup(ctx, paths, excludes, tags, handle)
finishedAt := time.Now().UTC()
status := api.JobSucceeded
exit := 0
errMsg := ""
if err != nil {
status = api.JobFailed
exit = -1
errMsg = err.Error()
}
var statsBlob json.RawMessage
if summary != nil {
statsBlob, _ = json.Marshal(summary)
}
finEnv, _ := api.Marshal(api.MsgJobFinished, jobID, api.JobFinishedPayload{
JobID: jobID,
Status: status,
ExitCode: exit,
FinishedAt: finishedAt,
Stats: statsBlob,
Error: errMsg,
})
_ = r.tx.Send(finEnv)
// On a successful backup, refresh the server's snapshot projection.
// We do this *after* job.finished so the UI sees the job land first;
// the snapshot list is a follow-up that the host detail page polls
// or the dashboard sees on its next refresh. A failure here is
// logged but doesn't fail the job — the next successful backup will
// catch the projection up.
if err == nil {
if rerr := r.reportSnapshots(ctx, env); rerr != nil {
slog.Warn("runner: snapshots.report failed", "job_id", jobID, "err", rerr)
}
}
if err != nil {
return fmt.Errorf("runner backup: %w", err)
}
return nil
}
// reportSnapshots calls `restic snapshots --json`, translates the
// payload into the wire shape, and ships it as a snapshots.report
// envelope. Bounded by a separate timeout so a sluggish repo doesn't
// hang the runner forever; restic snapshots is normally sub-second.
func (r *Runner) reportSnapshots(ctx context.Context, env restic.Env) error {
listCtx, cancel := context.WithTimeout(ctx, 60*time.Second)
defer cancel()
snaps, err := env.ListSnapshots(listCtx)
if err != nil {
return err
}
out := make([]api.Snapshot, len(snaps))
for i, s := range snaps {
out[i] = api.Snapshot{
ID: s.ID,
ShortID: s.ShortID,
Time: s.Time.UTC(),
Hostname: s.Hostname,
Paths: s.Paths,
Tags: s.Tags,
}
if s.Summary != nil {
out[i].SizeBytes = s.Summary.TotalBytesProcessed
out[i].FileCount = s.Summary.TotalFilesProcessed
}
}
envOut, err := api.Marshal(api.MsgSnapshotsRpt, "", api.SnapshotsReportPayload{
Snapshots: out,
})
if err != nil {
return err
}
return r.tx.Send(envOut)
}
func throughput(bytesDone, secondsElapsed int64) int64 {
if secondsElapsed <= 0 {
return 0
}
return bytesDone / secondsElapsed
}