minio/cmd/peer-s3-client.go
Harshavardhana 2f9e2147f5
allow quota enforcement to rely on older values (#17351)
PUT calls cannot afford to have large latency build-ups due
to contentious usage.json, or worse letting them fail with
some unexpected error, this can happen when this file is
concurrently being updated via scanner or it is being
healed during a disk replacement heal.

However, these are fairly quick in theory, stressed clusters
can quickly show visible latency this can add up leading to
invalid errors returned during PUT.

It is perhaps okay for us to relax this error return requirement
instead, make sure that we log that we are proceeding to take in
the requests while the quota is using an older value for the quota
enforcement. These things will reconcile themselves eventually,
via scanner making sure to overwrite the usage.json.

Bonus: make sure that storage-rest-client sets ExpectTimeouts to
be 'true', such that DiskInfo() call with contextTimeout does
not prematurely disconnect the servers leading to a longer
healthCheck, back-off routine. This can easily pile up while also
causing active callers to disconnect, leading to quorum loss.

DiskInfo is actively used in the PUT, Multipart call path for
upgrading parity when disks are down, it in-turn shouldn't cause
more disks to go down.
2023-06-05 16:56:35 -07:00

365 lines
10 KiB
Go

// Copyright (c) 2015-2022 MinIO, Inc.
//
// This file is part of MinIO Object Storage stack
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
package cmd
import (
"context"
"encoding/gob"
"errors"
"fmt"
"io"
"net/url"
"strconv"
xhttp "github.com/minio/minio/internal/http"
"github.com/minio/minio/internal/logger"
"github.com/minio/minio/internal/rest"
xnet "github.com/minio/pkg/net"
"github.com/minio/pkg/sync/errgroup"
)
var errPeerOffline = errors.New("peer is offline")
// client to talk to peer Nodes.
type peerS3Client struct {
host *xnet.Host
restClient *rest.Client
}
// Wrapper to restClient.Call to handle network errors, in case of network error the connection is marked disconnected
// permanently. The only way to restore the connection is at the xl-sets layer by xlsets.monitorAndConnectEndpoints()
// after verifying format.json
func (client *peerS3Client) call(method string, values url.Values, body io.Reader, length int64) (respBody io.ReadCloser, err error) {
return client.callWithContext(GlobalContext, method, values, body, length)
}
// Wrapper to restClient.Call to handle network errors, in case of network error the connection is marked disconnected
// permanently. The only way to restore the connection is at the xl-sets layer by xlsets.monitorAndConnectEndpoints()
// after verifying format.json
func (client *peerS3Client) callWithContext(ctx context.Context, method string, values url.Values, body io.Reader, length int64) (respBody io.ReadCloser, err error) {
if values == nil {
values = make(url.Values)
}
respBody, err = client.restClient.Call(ctx, method, values, body, length)
if err == nil {
return respBody, nil
}
err = toStorageErr(err)
return nil, err
}
// S3PeerSys - S3 peer call system.
type S3PeerSys struct {
peerClients []*peerS3Client // Excludes self
allPeerClients []*peerS3Client // Includes nil client for self
}
// NewS3PeerSys - creates new S3 peer calls.
func NewS3PeerSys(endpoints EndpointServerPools) *S3PeerSys {
remote, all := newPeerS3Clients(endpoints)
return &S3PeerSys{
peerClients: remote,
allPeerClients: all,
}
}
// ListBuckets lists buckets across all servers and returns a possible consistent view
func (sys *S3PeerSys) ListBuckets(ctx context.Context, opts BucketOptions) (result []BucketInfo, err error) {
g := errgroup.WithNErrs(len(sys.peerClients))
localBuckets, err := listBucketsLocal(ctx, opts)
if err != nil {
return nil, err
}
nodeBuckets := make([][]BucketInfo, len(sys.peerClients)+1)
errs := []error{nil}
nodeBuckets[0] = localBuckets
for idx, client := range sys.peerClients {
idx := idx
client := client
g.Go(func() error {
if client == nil {
return errPeerOffline
}
localBuckets, err := client.ListBuckets(ctx, opts)
if err != nil {
return err
}
nodeBuckets[idx+1] = localBuckets
return nil
}, idx)
}
errs = append(errs, g.Wait()...)
quorum := (len(sys.allPeerClients) / 2)
if err = reduceReadQuorumErrs(ctx, errs, bucketOpIgnoredErrs, quorum); err != nil {
return nil, err
}
bucketsMap := make(map[string]struct{})
for idx, buckets := range nodeBuckets {
if errs[idx] != nil {
continue
}
for _, bi := range buckets {
_, ok := bucketsMap[bi.Name]
if !ok {
bucketsMap[bi.Name] = struct{}{}
result = append(result, bi)
}
}
}
return result, nil
}
// GetBucketInfo returns bucket stat info about bucket on disk across all peers
func (sys *S3PeerSys) GetBucketInfo(ctx context.Context, bucket string, opts BucketOptions) (binfo BucketInfo, err error) {
g := errgroup.WithNErrs(len(sys.peerClients))
bucketInfos := make([]BucketInfo, len(sys.peerClients))
for idx, client := range sys.peerClients {
idx := idx
client := client
g.Go(func() error {
if client == nil {
return errPeerOffline
}
bucketInfo, err := client.GetBucketInfo(ctx, bucket, opts)
if err != nil {
return err
}
bucketInfos[idx] = bucketInfo
return nil
}, idx)
}
errs := g.Wait()
bucketInfo, err := getBucketInfoLocal(ctx, bucket, opts)
errs = append(errs, err)
bucketInfos = append(bucketInfos, bucketInfo)
quorum := (len(sys.allPeerClients) / 2)
if err = reduceReadQuorumErrs(ctx, errs, bucketOpIgnoredErrs, quorum); err != nil {
return BucketInfo{}, toObjectErr(err, bucket)
}
for i, err := range errs {
if err == nil {
return bucketInfos[i], nil
}
}
return BucketInfo{}, toObjectErr(errVolumeNotFound, bucket)
}
func (client *peerS3Client) ListBuckets(ctx context.Context, opts BucketOptions) ([]BucketInfo, error) {
v := url.Values{}
v.Set(peerS3BucketDeleted, strconv.FormatBool(opts.Deleted))
respBody, err := client.call(peerS3MethodListBuckets, v, nil, -1)
if err != nil {
return nil, err
}
defer xhttp.DrainBody(respBody)
var buckets []BucketInfo
err = gob.NewDecoder(respBody).Decode(&buckets)
return buckets, err
}
// GetBucketInfo returns bucket stat info from a peer
func (client *peerS3Client) GetBucketInfo(ctx context.Context, bucket string, opts BucketOptions) (BucketInfo, error) {
v := url.Values{}
v.Set(peerS3Bucket, bucket)
v.Set(peerS3BucketDeleted, strconv.FormatBool(opts.Deleted))
respBody, err := client.call(peerS3MethodGetBucketInfo, v, nil, -1)
if err != nil {
return BucketInfo{}, err
}
defer xhttp.DrainBody(respBody)
var bucketInfo BucketInfo
err = gob.NewDecoder(respBody).Decode(&bucketInfo)
return bucketInfo, err
}
// MakeBucket creates bucket across all peers
func (sys *S3PeerSys) MakeBucket(ctx context.Context, bucket string, opts MakeBucketOptions) error {
g := errgroup.WithNErrs(len(sys.peerClients))
for idx, client := range sys.peerClients {
client := client
g.Go(func() error {
if client == nil {
return errPeerOffline
}
return client.MakeBucket(ctx, bucket, opts)
}, idx)
}
errs := g.Wait()
errs = append(errs, makeBucketLocal(ctx, bucket, opts))
quorum := (len(sys.allPeerClients) / 2) + 1
err := reduceWriteQuorumErrs(ctx, errs, bucketOpIgnoredErrs, quorum)
// Perform MRF on missing buckets for temporary errors.
for _, err := range errs {
if err == nil {
continue
}
if errors.Is(err, errPeerOffline) || errors.Is(err, errDiskNotFound) ||
isNetworkError(err) {
globalMRFState.addPartialOp(partialOperation{
bucket: bucket,
})
}
}
return toObjectErr(err, bucket)
}
// MakeBucket creates a bucket on a peer
func (client *peerS3Client) MakeBucket(ctx context.Context, bucket string, opts MakeBucketOptions) error {
v := url.Values{}
v.Set(peerS3Bucket, bucket)
v.Set(peerS3BucketForceCreate, strconv.FormatBool(opts.ForceCreate))
respBody, err := client.call(peerS3MethodMakeBucket, v, nil, -1)
if err != nil {
return err
}
defer xhttp.DrainBody(respBody)
return nil
}
// DeleteBucket deletes bucket across all peers
func (sys *S3PeerSys) DeleteBucket(ctx context.Context, bucket string, opts DeleteBucketOptions) error {
g := errgroup.WithNErrs(len(sys.peerClients))
for idx, client := range sys.peerClients {
client := client
g.Go(func() error {
if client == nil {
return errPeerOffline
}
return client.DeleteBucket(ctx, bucket, opts)
}, idx)
}
errs := g.Wait()
errs = append(errs, deleteBucketLocal(ctx, bucket, opts))
var errReturn error
for _, err := range errs {
if errReturn == nil && err != nil {
// always return first error
errReturn = toObjectErr(err, bucket)
break
}
}
for _, err := range errs {
if err == nil && errReturn != nil {
// re-create successful deletes, since we are return an error.
sys.MakeBucket(ctx, bucket, MakeBucketOptions{})
break
}
}
return errReturn
}
// DeleteBucket deletes bucket on a peer
func (client *peerS3Client) DeleteBucket(ctx context.Context, bucket string, opts DeleteBucketOptions) error {
v := url.Values{}
v.Set(peerS3Bucket, bucket)
v.Set(peerS3BucketForceDelete, strconv.FormatBool(opts.Force))
respBody, err := client.call(peerS3MethodDeleteBucket, v, nil, -1)
if err != nil {
return err
}
defer xhttp.DrainBody(respBody)
return nil
}
// newPeerS3Clients creates new peer clients.
// The two slices will point to the same clients,
// but 'all' will contain nil entry for local client.
// The 'all' slice will be in the same order across the cluster.
func newPeerS3Clients(endpoints EndpointServerPools) (remote, all []*peerS3Client) {
if !globalIsDistErasure {
// Only useful in distributed setups
return nil, nil
}
hosts := endpoints.hostsSorted()
remote = make([]*peerS3Client, 0, len(hosts))
all = make([]*peerS3Client, len(hosts))
for i, host := range hosts {
if host == nil {
continue
}
all[i] = newPeerS3Client(host)
remote = append(remote, all[i])
}
if len(all) != len(remote)+1 {
logger.LogIf(context.Background(), fmt.Errorf("WARNING: Expected number of all hosts (%v) to be remote +1 (%v)", len(all), len(remote)))
}
return remote, all
}
// Returns a peer S3 client.
func newPeerS3Client(peer *xnet.Host) *peerS3Client {
scheme := "http"
if globalIsTLS {
scheme = "https"
}
serverURL := &url.URL{
Scheme: scheme,
Host: peer.String(),
Path: peerS3Path,
}
restClient := rest.NewClient(serverURL, globalInternodeTransport, newCachedAuthToken())
// Use a separate client to avoid recursive calls.
healthClient := rest.NewClient(serverURL, globalInternodeTransport, newCachedAuthToken())
healthClient.NoMetrics = true
// Construct a new health function.
restClient.HealthCheckFn = func() bool {
ctx, cancel := context.WithTimeout(context.Background(), restClient.HealthCheckTimeout)
defer cancel()
respBody, err := healthClient.Call(ctx, peerS3MethodHealth, nil, nil, -1)
xhttp.DrainBody(respBody)
return !isNetworkError(err)
}
return &peerS3Client{host: peer, restClient: restClient}
}