- **We do not recommend upgrading one MinIO server at a time, the product is designed to support parallel upgrades please follow our recommended guidelines.**

## Explore Further

- [MinIO Erasure Code Overview](https://docs.min.io/community/minio-object-store/operations/concepts/erasure-coding.html)

single process running those previous steps *even* if afterwards, you start **multiple processes** (multiple workers) for the application itself. If those steps were run by **multiple processes**, they would **duplicate** the work by running it in **parallel**, and if the steps were something delicate like a database migration, they could cause conflicts with each other.

Of course, there are some cases where there's no problem in running the previous steps multiple times, in that case, it's...

          lastGoodStreamId = streamId
          streams[streamId] = newStream

          // Use a different task queue for each stream because they should be handled in parallel.
          taskRunner.newQueue().execute("$connectionName[$streamId] onStream") {
            try {
              listener.onStream(newStream)
            } catch (e: IOException) {

    ChannelManager channelManager = session.getChannelManager();
    List<ChannelInfo> channels = channelManager.getHealthyChannels();
    int channelCount = Math.min(channels.size(), 4);  // Max 4 parallel reads
    
    int chunkSize = length / channelCount;
    List<CompletableFuture<Void>> futures = new ArrayList<>();
    
    for (int i = 0; i < channelCount; i++) {
        final int chunkOffset = i * chunkSize;

	bootstrapTraceMsgFirstTime("loading policy documents")

	policyLoadStartTime := UTCNow()
	policiesList := listedConfigItems[policiesListKey]
	count := 32 // number of parallel IAM loaders
	for {
		if len(policiesList) < count {
			policyDocs, err := iamOS.loadPolicyDocConcurrent(ctx, policiesList...)
			if err != nil {
				return err
			}
			for index := range policiesList {

	for _, lstorageInfo := range storageInfos {
		storageInfo.Disks = append(storageInfo.Disks, lstorageInfo.Disks...)
	}

	return storageInfo
}

// Shutdown shutsdown all erasure coded sets in parallel
// returns error upon first error.
func (s *erasureSets) Shutdown(ctx context.Context) error {
	g := errgroup.WithNErrs(len(s.sets))

	for index := range s.sets {
		g.Go(func() error {

///

Wenn es in Ihrem Anwendungsfall kein Problem darstellt, diese vorherigen Schritte **mehrmals parallel** auszuführen (z. B. wenn Sie keine Datenbankmigrationen ausführen, sondern nur prüfen, ob die Datenbank bereits bereit ist), können Sie sie auch einfach in jedem Container direkt vor dem Start des Hauptprozesses einfügen.

4 }} {{- end }} {{- if .Values.additionalAnnotation }} annotations: {{ toYaml .Values.additionalAnnotation | trimSuffix "\n" | indent 4 }} {{- end }} spec: updateStrategy: type: {{ .Values.StatefulSetUpdate.updateStrategy }} podManagementPolicy: "Parallel" serviceName: {{ template "minio.fullname" . }}-svc replicas: {{ mul $poolCount $nodeCount }} selector: matchLabels: app: {{ template "minio.name" . }} release: {{ .Release.Name }} template: metadata: name: {{ template "minio.fullname" . }} labels:...

				"multipart/byteranges",
				"multipart/digest",
				"multipart/encrypted",
				"multipart/example",
				"multipart/form-data",
				"multipart/header-set",
				"multipart/mixed",
				"multipart/parallel",
				"multipart/related",
				"multipart/report",
				"multipart/signed",
				"multipart/voice-message",
				"text/x-actionscript",
				"text/x-ada",
				"text/x-applescript",
				"text/asp",

///

If in your use case there's no problem in running those previous steps **multiple times in parallel** (for example if you are not running database migrations, but just checking if the database is ready yet), then you could also just put them in each container right before starting the main process.