Merge remote-tracking branch 'upstream/release-0.13' into

openshift/master

* upstream/release-0.13:
  Cut release v0.13.0
  shipper: Be strict about upload order unless it's specified so & cut v0.13.0-rc.2 (thanos-io#2765)
  Cut 0.13.0 release. (thanos-io#2762)
  Cut release 0.13.0-rc.1 (thanos-io#2720)
  Store: `irate` and `resets` use now counter downsampling aggregations. (thanos-io#2719)
  deps: Updated minio-go dependency to v6.0.56 to add two region endpoints (thanos-io#2705) (thanos-io#2718)
  store/proxy: Deduplicate chunks on StoreAPI level. Recommend chunk sorting for StoreAPI + Optimized iter chunk dedup. (thanos-io#2710) (thanos-io#2711)
  Allow using multiple memcached clients at the same time. (thanos-io#2648) (thanos-io#2698)
  Updated Prometheus as little as possible to include Isolation fix. (thanos-io#2697)
  Release fix attempt2.
  Fixed test job. (thanos-io#2650)
  Fixed promu build to build in compatible directory that crossbuild understands.
  Cut v0.13.0-rc.0 (thanos-io#2628)

.circleci/config.yml

@@ -4,7 +4,7 @@ version: 2
 defaults: &defaults
   docker:
     # Built by Thanos make docker-ci
-    - image: quay.io/thanos/thanos-ci:v1.2-go1.14.2-node
+    - image: quay.io/thanos/thanos-ci:v1.1-go1.14.2-node
 jobs:
   test:
     <<: *defaults

CHANGELOG.md

@@ -11,63 +11,48 @@ We use *breaking* word for marking changes that are not backward compatible (rel
 
 ## Unreleased
 
-### Fixed
-
-- [#2637](https://github.com/thanos-io/thanos/pull/2637) Compact: detect retryable errors that are inside of a wrapped `tsdb.MultiError`
-- [#2648](https://github.com/thanos-io/thanos/pull/2648) Store: allow index cache and caching bucket to be configured at the same time.
-- [#2705](https://github.com/thanos-io/thanos/pull/2705) minio-go: Added support for `af-south-1` and `eu-south-1` regions.
-- [#2728](https://github.com/thanos-io/thanos/pull/2728) Query: Fixed panics when using larger number of replica labels with short series label sets.
-
-### Changed
-
-- [#2658](https://github.com/thanos-io/thanos/pull/2658) [#2703](https://github.com/thanos-io/thanos/pull/2703) Upgrade to Prometheus [@3268eac2ddda](https://github.com/prometheus/prometheus/commit/3268eac2ddda) which is after v2.18.1.
-    - TSDB now does memory-mapping of Head chunks and reduces memory usage.
-- [#2667](https://github.com/thanos-io/thanos/pull/2667) Store: removed support to the legacy `index.cache.json`. The hidden flag `--store.disable-index-header` was removed.
-- [#2667](https://github.com/thanos-io/thanos/pull/2667) Compact: the deprecated flag `--index.generate-missing-cache-file` and the metric `thanos_compact_generated_index_total` were removed.
-- [2603](https://github.com/thanos-io/thanos/pull/2603) Store/Querier: Significantly optimize cases where StoreAPIs or blocks returns exact overlapping chunks (e.g Store GW and sidecar or brute force Store Gateway HA).
-- [#2671](https://github.com/thanos-io/thanos/pull/2671) *breaking* Tools: bucket replicate flag `--resolution` is now in Go duration format.
-- [#2671](https://github.com/thanos-io/thanos/pull/2671) Tools: bucket replicate now replicates by default all blocks.
-- [#2739](https://github.com/thanos-io/thanos/pull/2739) Changed `bucket tool bucket verify` `--id-whitelist` flag to `--id`.
-- [#2748](https://github.com/thanos-io/thanos/pull/2748) Upgrade Prometheus to [@66dfb951c4ca](https://github.com/prometheus/prometheus/commit/66dfb951c4ca2c1dd3f266172a48a925403b13a5) which is after v2.19.0.
-    - PromQL now allow us to executed concurrent selects.
-
-### Added
-
-- [#2671](https://github.com/thanos-io/thanos/pull/2671) Tools: bucket replicate now allows passing repeated `--compaction` and `--resolution` flags.
-
-## [v0.13.0](https://github.com/thanos-io/thanos/releases) - IN PROGRESS
+## [v0.13.0](https://github.com/thanos-io/thanos/releases/tag/v0.13.0) - 2020.06.22
 
 ### Fixed
 
-- [#2536](https://github.com/thanos-io/thanos/pull/2536) minio-go: Fixed AWS STS endpoint url to https for Web Identity providers on AWS EKS
+- [#2548](https://github.com/thanos-io/thanos/pull/2548) Query: Fixed rare cases of double counter reset accounting when querying `rate` with deduplication enabled.
+- [#2536](https://github.com/thanos-io/thanos/pull/2536) S3: Fixed AWS STS endpoint url to https for Web Identity providers on AWS EKS.
 - [#2501](https://github.com/thanos-io/thanos/pull/2501) Query: Gracefully handle additional fields in `SeriesResponse` protobuf message that may be added in the future.
-- [#2568](https://github.com/thanos-io/thanos/pull/2568) Query: Does not close the connection of strict, static nodes if establishing a connection had succeeded but Info() call failed
+- [#2568](https://github.com/thanos-io/thanos/pull/2568) Query: Don't close the connection of strict, static nodes if establishing a connection had succeeded but Info() call failed.
 - [#2615](https://github.com/thanos-io/thanos/pull/2615) Rule: Fix bugs where rules were out of sync.
-- [#2614](https://github.com/thanos-io/thanos/pull/2614) Tracing: Disable Elastic APM Go Agent default tracer on initialization to disable the default metric gatherer
-- [#2548](https://github.com/thanos-io/thanos/pull/2548) Query: Fixed rare cases of double counter reset accounting when querying `rate` with deduplication enabled.
+- [#2614](https://github.com/thanos-io/thanos/pull/2614) Tracing: Disabled Elastic APM Go Agent default tracer on initialization to disable the default metric gatherer.
 - [#2525](https://github.com/thanos-io/thanos/pull/2525) Query: Fixed logging for dns resolution error in the `Query` component.
+- [#2484](https://github.com/thanos-io/thanos/pull/2484) Query/Ruler: Fixed issue #2483, when web.route-prefix is set, it is added twice in HTTP router prefix.
+- [#2416](https://github.com/thanos-io/thanos/pull/2416) Bucket: Fixed issue #2416 bug in `inspect --sort-by` doesn't work correctly in all cases.
+- [#2719](https://github.com/thanos-io/thanos/pull/2719) Query: `irate` and `resets` use now counter downsampling aggregations.
+- [#2705](https://github.com/thanos-io/thanos/pull/2705) minio-go: Added support for `af-south-1` and `eu-south-1` regions.
+- [#2753](https://github.com/thanos-io/thanos/issues/2753) Sidecar, Receive, Rule: Fixed possibility of out of order uploads in error cases. This could potentially cause Compactor to create overlapping blocks.
 
 ### Added
 
-- [#2502](https://github.com/thanos-io/thanos/pull/2502) Added `hints` field to `SeriesResponse`. Hints in an opaque data structure that can be used to carry additional information from the store and its content is implementation specific.
-- [#2521](https://github.com/thanos-io/thanos/pull/2521) Sidecar: add `thanos_sidecar_reloader_reloads_failed_total`, `thanos_sidecar_reloader_reloads_total`, `thanos_sidecar_reloader_watch_errors_total`, `thanos_sidecar_reloader_watch_events_total` and `thanos_sidecar_reloader_watches` metrics.
-- [#2412](https://github.com/thanos-io/thanos/pull/2412) ui: add React UI from Prometheus upstream. Currently only accessible from Query component as only `/graph` endpoint is migrated.
+- [#2012](https://github.com/thanos-io/thanos/pull/2012) Receive: Added multi-tenancy support (based on header)
+- [#2502](https://github.com/thanos-io/thanos/pull/2502) StoreAPI: Added `hints` field to `SeriesResponse`. Hints in an opaque data structure that can be used to carry additional information from the store and its content is implementation specific.
+- [#2521](https://github.com/thanos-io/thanos/pull/2521) Sidecar: Added `thanos_sidecar_reloader_reloads_failed_total`, `thanos_sidecar_reloader_reloads_total`, `thanos_sidecar_reloader_watch_errors_total`, `thanos_sidecar_reloader_watch_events_total` and `thanos_sidecar_reloader_watches` metrics.
+- [#2412](https://github.com/thanos-io/thanos/pull/2412) UI: Added React UI from Prometheus upstream. Currently only accessible from Query component as only `/graph` endpoint is migrated.
 - [#2532](https://github.com/thanos-io/thanos/pull/2532) Store: Added hidden option `--store.caching-bucket.config=<yaml content>` (or `--store.caching-bucket.config-file=<file.yaml>`) for experimental caching bucket, that can cache chunks into shared memcached. This can speed up querying and reduce number of requests to object storage.
 - [#2579](https://github.com/thanos-io/thanos/pull/2579) Store: Experimental caching bucket can now cache metadata as well. Config has changed from #2532.
 - [#2526](https://github.com/thanos-io/thanos/pull/2526) Compact: In case there are no labels left after deduplication via `--deduplication.replica-label`, assign first `replica-label` with value `deduped`.
-- [#2621](https://github.com/thanos-io/thanos/pull/2621) Receive: add flag to configure forward request timeout.
+- [#2621](https://github.com/thanos-io/thanos/pull/2621) Receive: Added flag to configure forward request timeout. Receive write will complete request as soon as quorum of writes succeeds.
 
 ### Changed
 
-- [#2194](https://github.com/thanos-io/thanos/pull/2194) Updated to golang v1.14.2
-- [#2505](https://github.com/thanos.io/thanos/pull/2505) Store: remove obsolete `thanos_store_node_info` metric.
+- [#2194](https://github.com/thanos-io/thanos/pull/2194) Updated to golang v1.14.2.
+- [#2505](https://github.com/thanos.io/thanos/pull/2505) Store: Removed obsolete `thanos_store_node_info` metric.
 - [2513](https://github.com/thanos-io/thanos/pull/2513) Tools: Moved `thanos bucket` commands to `thanos tools bucket`, also
 moved `thanos check rules` to `thanos tools rules-check`. `thanos tools rules-check` also takes rules by `--rules` repeated flag not argument
 anymore.
 - [#2548](https://github.com/thanos-io/thanos/pull/2548/commits/53e69bd89b2b08c18df298eed7d90cb7179cc0ec) Store, Querier: remove duplicated chunks on StoreAPI.
-- [#2596](https://github.com/thanos-io/thanos/pull/2596) Update to Prometheus [@cd73b3d33e064bbd846fc7a26dc8c313d46af382](https://github.com/prometheus/prometheus/commit/cd73b3d33e064bbd846fc7a26dc8c313d46af382) which falls in between v2.17.0 and v2.18.0.
-    - TSDB now supports isolation of append and queries.
-    - TSDB now holds less WAL files after Head Truncation.
+- [#2596](https://github.com/thanos-io/thanos/pull/2596) Updated Prometheus dependency to [@cd73b3d33e064bbd846fc7a26dc8c313d46af382](https://github.com/prometheus/prometheus/commit/cd73b3d33e064bbd846fc7a26dc8c313d46af382) which falls in between v2.17.0 and v2.18.0.
+    - Receive,Rule: TSDB now supports isolation of append and queries.
+    - Receive,Rule: TSDB now holds less WAL files after Head Truncation.
+- [#2450](https://github.com/thanos-io/thanos/pull/2450) Store: Added Regex-set optimization for `label=~"a|b|c"` matchers.
+- [#2526](https://github.com/thanos-io/thanos/pull/2526) Compact: In case there are no labels left after deduplication via `--deduplication.replica-label`, assign first `replica-label` with value `deduped`.
+- [2603](https://github.com/thanos-io/thanos/pull/2603) Store/Querier: Significantly optimize cases where StoreAPIs or blocks returns exact overlapping chunks (e.g Store GW and sidecar or brute force Store Gateway HA).
 
 ## [v0.12.2](https://github.com/thanos-io/thanos/releases/tag/v0.12.2) - 2020.04.30
 
@@ -81,17 +66,9 @@ anymore.
 ### Fixed
 
 - [#2411](https://github.com/thanos-io/thanos/pull/2411) Query: fix a bug where queries might not time out sometimes due to issues with one or more StoreAPIs.
-- [#2474](https://github.com/thanos-io/thanos/pull/2474) Store: fix a panic caused by concurrent memory access during block filtering.
-- [#2472](https://github.com/thanos-io/thanos/pull/2472) Compact: fix a bug where partial blocks were never deleted, causing spam of warnings.
-- [#2484](https://github.com/thanos-io/thanos/pull/2484) Query/Ruler: fix issue #2483, when web.route-prefix is set, it is added twice in HTTP router prefix.
-- [#2416](https://github.com/thanos-io/thanos/pull/2416) Bucket: fixes issue #2416 bug in `inspect --sort-by` doesn't work correctly in all cases
-- [#2411](https://github.com/thanos-io/thanos/pull/2411) Query: fix a bug where queries might not time out sometimes due to issues with one or more StoreAPIs
-
-### Added
-
-### Changed
-- [#2450](https://github.com/thanos-io/thanos/pull/2450) Store: regex-set optimization for `label=~"a|b|c"` matchers.
 - [#2475](https://github.com/thanos-io/thanos/pull/2475) Store: remove incorrect optimizations for queries with `=~".*"` and `!=~".*"` matchers.
+- [#2472](https://github.com/thanos-io/thanos/pull/2472) Compact: fix a bug where partial blocks were never deleted, causing spam of warnings.
+- [#2474](https://github.com/thanos-io/thanos/pull/2474) Store: fix a panic caused by concurrent memory access during block filtering.
 
 ## [v0.12.0](https://github.com/thanos-io/thanos/releases/tag/v0.12.0) - 2020.04.15
 

Dockerfile.thanos-ci

@@ -1,6 +1,6 @@
 # Available from https://hub.docker.com/r/circleci/golang/
 FROM circleci/golang:1.14.2-node
 
-ENV GOBIN=/go/bin
+ENV GOBIN=/home/circleci/.local/bin
 
-COPY tmp/bin ${GOBIN}
+COPY --chown=circleci tmp/bin/* ${GOBIN}/

Makefile

@@ -11,6 +11,10 @@ GOPATH            ?= $(shell go env GOPATH)
 
 TMP_GOPATH        ?= /tmp/thanos-go
 GOBIN             ?= $(firstword $(subst :, ,${GOPATH}))/bin
+
+# Promu is using this exact variable name, do not rename.
+PREFIX  ?= $(GOBIN)
+
 GO111MODULE       ?= on
 export GO111MODULE
 GOPROXY           ?= https://proxy.golang.org
@@ -104,8 +108,8 @@ react-app-start: $(REACT_APP_NODE_MODULES_PATH)
 .PHONY: build
 build: ## Builds Thanos binary using `promu`.
 build: check-git deps $(PROMU)
-	@echo ">> building Thanos binary in $(GOBIN)"
-	@$(PROMU) build --prefix $(GOBIN)
+	@echo ">> building Thanos binary in $(PREFIX)"
+	@$(PROMU) build --prefix $(PREFIX)
 
 .PHONY: crossbuild
 crossbuild: ## Builds all binaries for all platforms.
@@ -121,8 +125,8 @@ deps: ## Ensures fresh go.mod and go.sum.
 .PHONY: docker
 docker: ## Builds 'thanos' docker with no tag.
 docker: build
-	@echo ">> copying Thanos from $(GOBIN) to ./thanos_tmp_for_docker"
-	@cp $(GOBIN)/thanos ./thanos_tmp_for_docker
+	@echo ">> copying Thanos from $(PREFIX) to ./thanos_tmp_for_docker"
+	@cp $(PREFIX)/thanos ./thanos_tmp_for_docker
 	@echo ">> building docker image 'thanos'"
 	@docker build -t "thanos" .
 	@rm ./thanos_tmp_for_docker

VERSION

@@ -1 +1 @@
-0.12.2
+0.13.0

cmd/thanos/config.go

@@ -112,17 +112,23 @@ func (rc *reloaderConfig) registerFlag(cmd *kingpin.CmdClause) *reloaderConfig {
 }
 
 type shipperConfig struct {
-	uploadCompacted bool
-	ignoreBlockSize bool
+	uploadCompacted       bool
+	ignoreBlockSize       bool
+	allowOutOfOrderUpload bool
 }
 
 func (sc *shipperConfig) registerFlag(cmd *kingpin.CmdClause) *shipperConfig {
 	cmd.Flag("shipper.upload-compacted",
-		"If true sidecar will try to upload compacted blocks as well. Useful for migration purposes. Works only if compaction is disabled on Prometheus. Do it once and then disable the flag when done.").
+		"If true shipper will try to upload compacted blocks as well. Useful for migration purposes. Works only if compaction is disabled on Prometheus. Do it once and then disable the flag when done.").
 		Default("false").BoolVar(&sc.uploadCompacted)
 	cmd.Flag("shipper.ignore-unequal-block-size",
-		"If true sidecar will not require prometheus min and max block size flags to be set to the same value. Only use this if you want to keep long retention and compaction enabled on your Prometheus instance, as in the worst case it can result in ~2h data loss for your Thanos bucket storage.").
+		"If true shipper will not require prometheus min and max block size flags to be set to the same value. Only use this if you want to keep long retention and compaction enabled on your Prometheus instance, as in the worst case it can result in ~2h data loss for your Thanos bucket storage.").
 		Default("false").Hidden().BoolVar(&sc.ignoreBlockSize)
+	cmd.Flag("shipper.allow-out-of-order-uploads",
+		"If true, shipper will skip failed block uploads in the given iteration and retry later. This means that some newer blocks might be uploaded sooner than older blocks."+
+			"This can trigger compaction without those blocks and as a result will create an overlap situation. Set it to true if you have vertical compaction enabled and wish to upload blocks as soon as possible without caring"+
+			"about order.").
+		Default("false").Hidden().BoolVar(&sc.allowOutOfOrderUpload)
 	return sc
 }
 

cmd/thanos/receive.go

@@ -89,6 +89,12 @@ func registerReceive(m map[string]setupFunc, app *kingpin.Application) {
 
 	walCompression := cmd.Flag("tsdb.wal-compression", "Compress the tsdb WAL.").Default("true").Bool()
 
+	allowOutOfOrderUpload := cmd.Flag("shipper.allow-out-of-order-uploads",
+		"If true, shipper will skip failed block uploads in the given iteration and retry later. This means that some newer blocks might be uploaded sooner than older blocks."+
+			"This can trigger compaction without those blocks and as a result will create an overlap situation. Set it to true if you have vertical compaction enabled and wish to upload blocks as soon as possible without caring"+
+			"about order.").
+		Default("false").Hidden().Bool()
+
 	m[comp.String()] = func(g *run.Group, logger log.Logger, reg *prometheus.Registry, tracer opentracing.Tracer, _ <-chan struct{}, _ bool) error {
 		lset, err := parseFlagLabels(*labelStrs)
 		if err != nil {
@@ -157,6 +163,7 @@ func registerReceive(m map[string]setupFunc, app *kingpin.Application) {
 			*replicationFactor,
 			time.Duration(*forwardTimeout),
 			comp,
+			*allowOutOfOrderUpload,
 		)
 	}
 }
@@ -195,6 +202,7 @@ func runReceive(
 	replicationFactor uint64,
 	forwardTimeout time.Duration,
 	comp component.SourceStoreAPI,
+	allowOutOfOrderUpload bool,
 ) error {
 	logger = log.With(logger, "component", "receive")
 	level.Warn(logger).Log("msg", "setting up receive; the Thanos receive component is EXPERIMENTAL, it may break significantly without notice")
@@ -246,6 +254,7 @@ func runReceive(
 		lset,
 		tenantLabelName,
 		bkt,
+		allowOutOfOrderUpload,
 	)
 	writer := receive.NewWriter(log.With(logger, "component", "receive-writer"), dbs)
 	webHandler := receive.NewHandler(log.With(logger, "component", "receive-handler"), &receive.Options{

cmd/thanos/rule.go

@@ -114,6 +114,12 @@ func registerRule(m map[string]setupFunc, app *kingpin.Application) {
 	dnsSDResolver := cmd.Flag("query.sd-dns-resolver", "Resolver to use. Possible options: [golang, miekgdns]").
 		Default("golang").Hidden().String()
 
+	allowOutOfOrderUpload := cmd.Flag("shipper.allow-out-of-order-uploads",
+		"If true, shipper will skip failed block uploads in the given iteration and retry later. This means that some newer blocks might be uploaded sooner than older blocks."+
+			"This can trigger compaction without those blocks and as a result will create an overlap situation. Set it to true if you have vertical compaction enabled and wish to upload blocks as soon as possible without caring"+
+			"about order.").
+		Default("false").Hidden().Bool()
+
 	m[comp.String()] = func(g *run.Group, logger log.Logger, reg *prometheus.Registry, tracer opentracing.Tracer, reload <-chan struct{}, _ bool) error {
 		lset, err := parseFlagLabels(*labelStrs)
 		if err != nil {
@@ -197,6 +203,7 @@ func registerRule(m map[string]setupFunc, app *kingpin.Application) {
 			time.Duration(*dnsSDInterval),
 			*dnsSDResolver,
 			comp,
+			*allowOutOfOrderUpload,
 		)
 	}
 }
@@ -283,6 +290,7 @@ func runRule(
 	dnsSDInterval time.Duration,
 	dnsSDResolver string,
 	comp component.Component,
+	allowOutOfOrderUpload bool,
 ) error {
 	metrics := newRuleMetrics(reg)
 
@@ -610,7 +618,7 @@ func runRule(
 			}
 		}()
 
-		s := shipper.New(logger, reg, dataDir, bkt, func() labels.Labels { return lset }, metadata.RulerSource)
+		s := shipper.New(logger, reg, dataDir, bkt, func() labels.Labels { return lset }, metadata.RulerSource, allowOutOfOrderUpload)
 
 		ctx, cancel := context.WithCancel(context.Background())
 

cmd/thanos/sidecar.go

@@ -275,9 +275,9 @@ func runSidecar(
 
 			var s *shipper.Shipper
 			if conf.shipper.uploadCompacted {
-				s = shipper.NewWithCompacted(logger, reg, conf.tsdb.path, bkt, m.Labels, metadata.SidecarSource)
+				s = shipper.NewWithCompacted(logger, reg, conf.tsdb.path, bkt, m.Labels, metadata.SidecarSource, conf.shipper.allowOutOfOrderUpload)
 			} else {
-				s = shipper.New(logger, reg, conf.tsdb.path, bkt, m.Labels, metadata.SidecarSource)
+				s = shipper.New(logger, reg, conf.tsdb.path, bkt, m.Labels, metadata.SidecarSource, conf.shipper.allowOutOfOrderUpload)
 			}
 
 			return runutil.Repeat(30*time.Second, ctx.Done(), func() error {

docs/components/sidecar.md

@@ -151,7 +151,7 @@ Flags:
                                  details:
                                  https://thanos.io/storage.md/#configuration
       --shipper.upload-compacted
-                                 If true sidecar will try to upload compacted
+                                 If true shipper will try to upload compacted
                                  blocks as well. Useful for migration purposes.
                                  Works only if compaction is disabled on
                                  Prometheus. Do it once and then disable the