Update docs/victoriametrics/changelog/CHANGELOG.md

Co-authored-by: cubic-dev-ai[bot] <191113872+cubic-dev-ai[bot]@users.noreply.github.com>
Signed-off-by: JAYICE <jayice.zhou@qq.com>

app/vmagent/remotewrite/remotewrite.go

@@ -102,6 +102,8 @@ var (
 		"cannot be pushed into the configured -remoteWrite.url systems in a timely manner. See https://docs.victoriametrics.com/victoriametrics/vmagent/#disabling-on-disk-persistence")
 	disableMetadataPerURL = flagutil.NewArrayBool("remoteWrite.disableMetadata", "Whether to disable sending metadata to the corresponding -remoteWrite.url. "+
 		"By default, metadata sending is controlled by the global -enableMetadata flag")
+	enableRerouting = flag.Bool("remoteWrite.enableRerouting", false, "Whether to reroute samples to available remote storage systems when there's any remote storage system and its persistent queue can not "+
+		"keep up with the data ingestion rate. If this flag is not set, then it will be calculated automatically based on -remoteWrite.disableOnDiskQueue. See https://docs.victoriametrics.com/victoriametrics/vmagent/#disabling-on-disk-persistence")
 )
 
 var (
@@ -215,6 +217,10 @@ func Init() {
 	// to the remaining -remoteWrite.url and dropping them on the blocked queue.
 	dropSamplesOnFailureGlobal = *dropSamplesOnOverload || disableOnDiskQueueAny && len(*remoteWriteURLs) > 1
 
+	if *shardByURL && !flagutil.IsSet("remoteWrite.enableRerouting") {
+		*enableRerouting = disableOnDiskQueueAny
+	}
+
 	dropDanglingQueues()
 
 	// Start config reloader.
@@ -498,11 +504,13 @@ func tryPush(at *auth.Token, wr *prompb.WriteRequest, forceDropSamplesOnFailure
 //
 // calculateHealthyRwctxIdx will rely on the order of rwctx to be in ascending order.
 func getEligibleRemoteWriteCtxs(tss []prompb.TimeSeries, forceDropSamplesOnFailure bool) ([]*remoteWriteCtx, bool) {
-	if !disableOnDiskQueueAny {
+	if (*shardByURL && !*enableRerouting) || !disableOnDiskQueueAny {
 		return rwctxsGlobal, true
 	}
 
-	// This code is applicable if at least a single remote storage has -disableOnDiskQueue
+	// This code is applicable when:
+	// 1. remoteWrite.shardByUrl is disabled and at least a single remote storage has -disableOnDiskQueue.
+	// 2. remoteWrite.shardByUrl is enabled and remoteWrite.enableRerouting is set to true.
 	rwctxs := make([]*remoteWriteCtx, 0, len(rwctxsGlobal))
 	for _, rwctx := range rwctxsGlobal {
 		if !rwctx.fq.IsWriteBlocked() {

docs/victoriametrics/changelog/CHANGELOG.md

@@ -25,6 +25,7 @@ The sandbox cluster installation runs under the constant load generated by
 See also [LTS releases](https://docs.victoriametrics.com/victoriametrics/lts-releases/).
 
 ## tip
+* FEATURE: [vmagent](https://docs.victoriametrics.com/victoriametrics/vmagent/): introduce a new flag `-remoteWrite.enableRerouting` to explicitly control rerouting behavior when `vmagent` is started with `-remoteWrite.shardByUrl` and any remote storage system cannot keep up with the ingestion rate. Previously, this behavior was defined internally based on the value of `-remoteWrite.disableOnDiskQueue`. See [#10507](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10507).
 
 * FEATURE: all VictoriaMetrics components: add support for reading cpu/memory limits configured via [systemd slices](https://www.freedesktop.org/software/systemd/man/latest/systemd.slice.html). Previously, only limits set directly on the process's own cgroup were detected. See [#10635](https://github.com/VictoriaMetrics/VictoriaMetrics/issues/10635). Thanks to @andriibeee for the contribution.
 * FEATURE: [vmui](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#vmui): now `Run query` link on the Alerting Rules page correctly propagates the alert’s interval and evaluation time. See [#10366](https://github.com/VictoriaMetrics/VictoriaMetrics/pull/10366).

docs/victoriametrics/vmagent.md

@@ -900,6 +900,10 @@ When `-remoteWrite.disableOnDiskQueue` command-line flag is set, `vmagent` may s
 if it cannot keep up with the data ingestion rate. In this case [deduplication](https://docs.victoriametrics.com/victoriametrics/single-server-victoriametrics/#deduplication)
 must be enabled on all the configured remote storage systems.
 
+If `-remoteWrite.shardByUrl` command-line flag is set and `-remoteWrite.disableOnDiskQueue` is set for at least one `-remoteWrite.url`, then if any remote storage systems
+and its corresponding persistent queue can not keep up with the ingestion rate, `vmagent` will reroute samples to other available remote storage systems. 
+If on-disk persistence is enabled for all `-remoteWrite.url`, then `vmagent` will not perform rerouting. But the rerouting behavior can be controlled by explicitly setting `-remoteWrite.enableRerouting`.
+
 ## Cardinality limiter
 
 By default, `vmagent` doesn't limit the number of time series each scrape target can expose.

lib/flagutil/flag.go

@@ -18,3 +18,17 @@ func WriteFlags(w io.Writer) {
 		fmt.Fprintf(w, "-%s=%q\n", f.Name, value)
 	})
 }
+
+func IsSet(flagName string) bool {
+	isSet := false
+	flag.Visit(func(f *flag.Flag) {
+		if isSet {
+			return
+		}
+		lname := strings.ToLower(f.Name)
+		if flagName = strings.ToLower(flagName); lname == flagName {
+			isSet = true
+		}
+	})
+	return isSet
+}