From 4247168a2dae35648fa108f8f98b016a2843431e Mon Sep 17 00:00:00 2001
From: Roman Khavronenko <hagen1778@gmail.com>
Date: Mon, 10 May 2021 09:06:31 +0100
Subject: [PATCH] vmalert: fix error when rule didn't start if restore failed
 (#1279)

Previously, `startGroup` could exit on restore errors despite the
`remoteRead.ignoreRestoreErrors` flag value. Now vmalert checks the
flag value before deciding whether to return error or just log it.
---
 app/vmalert/alerting.go |  6 +-----
 app/vmalert/manager.go  | 13 +++++--------
 2 files changed, 6 insertions(+), 13 deletions(-)

diff --git a/app/vmalert/alerting.go b/app/vmalert/alerting.go
index 82a53348b..75c80f2de 100644
--- a/app/vmalert/alerting.go
+++ b/app/vmalert/alerting.go
@@ -2,7 +2,6 @@ package main
 
 import (
 	"context"
-	"errors"
 	"fmt"
 	"hash/fnv"
 	"sort"
@@ -405,9 +404,6 @@ func alertForToTimeSeries(name string, a *notifier.Alert, timestamp time.Time) p
 	return newTimeSeries(float64(a.Start.Unix()), labels, timestamp)
 }
 
-// ErrStateRestore indicates that the vmalert state failed to restore during startup.
-var ErrStateRestore = errors.New("failed to restore the state")
-
 // Restore restores the state of active alerts basing on previously written timeseries.
 // Restore restores only Start field. Field State will be always Pending and supposed
 // to be updated on next Exec, as well as Value field.
@@ -432,7 +428,7 @@ func (ar *AlertingRule) Restore(ctx context.Context, q datasource.Querier, lookb
 		alertForStateMetricName, ar.Name, labelsFilter, int(lookback.Seconds()))
 	qMetrics, err := q.Query(ctx, expr)
 	if err != nil {
-		return fmt.Errorf("%s: %w", err, ErrStateRestore)
+		return err
 	}
 
 	for _, m := range qMetrics {
diff --git a/app/vmalert/manager.go b/app/vmalert/manager.go
index 22de88ae8..4d9b0595e 100644
--- a/app/vmalert/manager.go
+++ b/app/vmalert/manager.go
@@ -2,7 +2,6 @@ package main
 
 import (
 	"context"
-	"errors"
 	"fmt"
 	"strings"
 	"sync"
@@ -52,12 +51,7 @@ func (m *manager) AlertAPI(gID, aID uint64) (*APIAlert, error) {
 }
 
 func (m *manager) start(ctx context.Context, path []string, validateTpl, validateExpr bool) error {
-	err := m.update(ctx, path, validateTpl, validateExpr, true)
-	if *remoteReadIgnoreRestoreErrors && errors.Is(err, ErrStateRestore) {
-		logger.Errorf("%s", err)
-		return nil
-	}
-	return err
+	return m.update(ctx, path, validateTpl, validateExpr, true)
 }
 
 func (m *manager) close() {
@@ -74,7 +68,10 @@ func (m *manager) startGroup(ctx context.Context, group *Group, restore bool) er
 	if restore && m.rr != nil {
 		err := group.Restore(ctx, m.rr, *remoteReadLookBack, m.labels)
 		if err != nil {
-			return fmt.Errorf("error while restoring state for group %q: %w", group.Name, err)
+			if !*remoteReadIgnoreRestoreErrors {
+				return fmt.Errorf("failed to restore state for group %q: %w", group.Name, err)
+			}
+			logger.Errorf("error while restoring state for group %q: %s", group.Name, err)
 		}
 	}