Use more of the cortex frontend module to deduplicate some code and add the memcached backend

CHANGELOG.md

@@ -11,6 +11,8 @@ We use *breaking :warning:* to mark changes that are not backward compatible (re
 
 ## Unreleased
 
+- [#3032](https://github.com/thanos-io/thanos/pull/3032) Query Frontend: Added support for Memacahce cache. Replaced underscores with hyphens in `log_queries_longer_than - > log-queries-longer-than`.
+
 ### Added
 
 - [#3133](https://github.com/thanos-io/thanos/pull/3133) Query: Allow passing a `storeMatch[]` to Labels APIs. Also time range metadata based store filtering is supported on Labels APIs.

cmd/thanos/query-frontend.go

@@ -8,101 +8,80 @@ import (
 	"time"
 
 	"github.com/NYTimes/gziphandler"
-	"github.com/cortexproject/cortex/pkg/querier/frontend"
+	cortexfrontend "github.com/cortexproject/cortex/pkg/querier/frontend"
 	"github.com/cortexproject/cortex/pkg/querier/queryrange"
+	cortexvalidation "github.com/cortexproject/cortex/pkg/util/validation"
 	"github.com/go-kit/kit/log"
 	"github.com/go-kit/kit/log/level"
 	"github.com/oklog/run"
 	"github.com/opentracing/opentracing-go"
 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
-	"github.com/prometheus/common/model"
-	"github.com/thanos-io/thanos/pkg/extkingpin"
-	"github.com/weaveworks/common/user"
-
 	"github.com/thanos-io/thanos/pkg/component"
 	"github.com/thanos-io/thanos/pkg/extflag"
+	"github.com/thanos-io/thanos/pkg/extkingpin"
 	"github.com/thanos-io/thanos/pkg/extprom"
 	extpromhttp "github.com/thanos-io/thanos/pkg/extprom/http"
 	"github.com/thanos-io/thanos/pkg/logging"
 	"github.com/thanos-io/thanos/pkg/prober"
 	"github.com/thanos-io/thanos/pkg/queryfrontend"
-	"github.com/thanos-io/thanos/pkg/queryfrontend/cache"
 	httpserver "github.com/thanos-io/thanos/pkg/server/http"
 	"github.com/thanos-io/thanos/pkg/server/http/middleware"
 	"github.com/thanos-io/thanos/pkg/tracing"
+	"github.com/weaveworks/common/user"
 )
 
 type queryFrontendConfig struct {
-	http             httpConfig
-	queryRangeConfig queryRangeConfig
-
-	downstreamURL        string
-	compressResponses    bool
-	LogQueriesLongerThan time.Duration
-
-	requestLoggingDecision string
+	http httpConfig
+	queryfrontend.Config
 }
 
-type queryRangeConfig struct {
-	respCacheConfig     extflag.PathOrContent
-	cacheMaxFreshness   time.Duration
-	splitInterval       model.Duration
-	maxRetries          int
-	maxQueryParallelism int
-	maxQueryLength      model.Duration
-
-	// partialResponseStrategy is the default strategy used
-	// when parsing thanos query request.
-	partialResponseStrategy bool
-}
+func registerQueryFrontend(app *extkingpin.App) {
+	comp := component.QueryFrontend
+	cmd := app.Command(comp.String(), "query frontend")
+	cfg := &queryFrontendConfig{
+		Config: queryfrontend.Config{
+			CortexFrontendConfig:     &cortexfrontend.Config{},
+			CortexLimits:             &cortexvalidation.Limits{},
+			CortexResultsCacheConfig: &queryrange.ResultsCacheConfig{},
+		},
+	}
 
-func (c *queryRangeConfig) registerFlag(cmd extkingpin.FlagClause) {
-	cmd.Flag("query-range.split-interval", "Split queries by an interval and execute in parallel, 0 disables it.").
-		Default("24h").SetValue(&c.splitInterval)
+	cfg.http.registerFlag(cmd)
+
+	cmd.Flag("query-range.split-interval", "Split queries by an interval and execute in parallel, it should be greater than 0 when response-cache-config is configured.").
+		Default("24h").DurationVar(&cfg.SplitQueriesByInterval)
 
 	cmd.Flag("query-range.max-retries-per-request", "Maximum number of retries for a single request; beyond this, the downstream error is returned.").
-		Default("5").IntVar(&c.maxRetries)
+		Default("5").IntVar(&cfg.MaxRetries)
 
 	cmd.Flag("query-range.max-query-length", "Limit the query time range (end - start time) in the query-frontend, 0 disables it.").
-		Default("0").SetValue(&c.maxQueryLength)
+		Default("0").DurationVar(&cfg.CortexLimits.MaxQueryLength)
 
-	cmd.Flag("query-range.max-query-parallelism", "Maximum number of queries will be scheduled in parallel by the frontend.").
-		Default("14").IntVar(&c.maxQueryParallelism)
+	cmd.Flag("query-range.max-query-parallelism", "Maximum number of queries will be scheduled in parallel by the Frontend.").
+		Default("14").IntVar(&cfg.CortexLimits.MaxQueryParallelism)
 
 	cmd.Flag("query-range.response-cache-max-freshness", "Most recent allowed cacheable result, to prevent caching very recent results that might still be in flux.").
-		Default("1m").DurationVar(&c.cacheMaxFreshness)
+		Default("1m").DurationVar(&cfg.CortexLimits.MaxCacheFreshness)
 
 	cmd.Flag("query-range.partial-response", "Enable partial response for queries if no partial_response param is specified. --no-query-range.partial-response for disabling.").
-		Default("true").BoolVar(&c.partialResponseStrategy)
+		Default("true").BoolVar(&cfg.PartialResponseStrategy)
 
-	c.respCacheConfig = *extflag.RegisterPathOrContent(cmd, "query-range.response-cache-config", "YAML file that contains response cache configuration.", false)
-}
-
-func (c *queryFrontendConfig) registerFlag(cmd extkingpin.FlagClause) {
-	c.queryRangeConfig.registerFlag(cmd)
-	c.http.registerFlag(cmd)
+	cfg.CachePathOrContent = *extflag.RegisterPathOrContent(cmd, "query-range.response-cache-config", "YAML file that contains response cache configuration.", false)
 
 	cmd.Flag("query-frontend.downstream-url", "URL of downstream Prometheus Query compatible API.").
-		Default("http://localhost:9090").StringVar(&c.downstreamURL)
+		Default("http://localhost:9090").StringVar(&cfg.CortexFrontendConfig.DownstreamURL)
 
 	cmd.Flag("query-frontend.compress-responses", "Compress HTTP responses.").
-		Default("false").BoolVar(&c.compressResponses)
-
-	cmd.Flag("query-frontend.log_queries_longer_than", "Log queries that are slower than the specified duration. "+
-		"Set to 0 to disable. Set to < 0 to enable on all queries.").Default("0").DurationVar(&c.LogQueriesLongerThan)
+		Default("false").BoolVar(&cfg.CortexFrontendConfig.CompressResponses)
 
-	cmd.Flag("log.request.decision", "Request Logging for logging the start and end of requests. LogFinishCall is enabled by default. LogFinishCall : Logs the finish call of the requests. LogStartAndFinishCall : Logs the start and finish call of the requests. NoLogCall : Disable request logging.").Default("LogFinishCall").EnumVar(&c.requestLoggingDecision, "NoLogCall", "LogFinishCall", "LogStartAndFinishCall")
-}
+	cmd.Flag("query-frontend.log-queries-longer-than", "Log queries that are slower than the specified duration. "+
+		"Set to 0 to disable. Set to < 0 to enable on all queries.").Default("0").DurationVar(&cfg.CortexFrontendConfig.LogQueriesLongerThan)
 
-func registerQueryFrontend(app *extkingpin.App) {
-	comp := component.QueryFrontend
-	cmd := app.Command(comp.String(), "query frontend")
-	conf := &queryFrontendConfig{}
-	conf.registerFlag(cmd)
+	cmd.Flag("log.request.decision", "Request Logging for logging the start and end of requests. LogFinishCall is enabled by default. LogFinishCall : Logs the finish call of the requests. LogStartAndFinishCall : Logs the start and finish call of the requests. NoLogCall : Disable request logging.").Default("LogFinishCall").EnumVar(&cfg.RequestLoggingDecision, "NoLogCall", "LogFinishCall", "LogStartAndFinishCall")
 
 	cmd.Setup(func(g *run.Group, logger log.Logger, reg *prometheus.Registry, tracer opentracing.Tracer, _ <-chan struct{}, _ bool) error {
-		return runQueryFrontend(g, logger, reg, tracer, conf, comp)
+		return runQueryFrontend(g, logger, reg, tracer, cfg, comp)
 	})
 }
 
@@ -111,54 +90,36 @@ func runQueryFrontend(
 	logger log.Logger,
 	reg *prometheus.Registry,
 	tracer opentracing.Tracer,
-	conf *queryFrontendConfig,
+	cfg *queryFrontendConfig,
 	comp component.Component,
 ) error {
+	cacheConfContentYaml, err := cfg.CachePathOrContent.Content()
+	if err != nil {
+		return err
+	}
+	if len(cacheConfContentYaml) > 0 {
+		cacheConfig, err := queryfrontend.NewCacheConfig(logger, cacheConfContentYaml)
+		if err != nil {
+			return errors.Wrap(err, "initializing the query frontend config")
+		}
+		if cfg.CortexResultsCacheConfig.CacheConfig.Memcache.Expiration == 0 {
+			level.Warn(logger).Log("msg", "memcached cache valid time set to 0, so using a default of 24 hours expiration time")
+			cfg.CortexResultsCacheConfig.CacheConfig.Memcache.Expiration = 24 * time.Hour
+		}
+		cfg.CortexResultsCacheConfig = cacheConfig
+	}
 
-	if len(conf.downstreamURL) == 0 {
-		return errors.New("downstream URL should be configured")
+	if err := cfg.Validate(); err != nil {
+		return errors.Wrap(err, "error validating the config")
 	}
 
-	fe, err := frontend.New(frontend.Config{
-		DownstreamURL:        conf.downstreamURL,
-		CompressResponses:    conf.compressResponses,
-		LogQueriesLongerThan: conf.LogQueriesLongerThan,
-	}, logger, reg)
+	fe, err := cortexfrontend.New(*cfg.CortexFrontendConfig, logger, reg)
 	if err != nil {
 		return errors.Wrap(err, "setup query frontend")
 	}
 	defer fe.Close()
 
-	limits := queryfrontend.NewLimits(
-		conf.queryRangeConfig.maxQueryParallelism,
-		time.Duration(conf.queryRangeConfig.maxQueryLength),
-		conf.queryRangeConfig.cacheMaxFreshness,
-	)
-
-	respCacheContentYaml, err := conf.queryRangeConfig.respCacheConfig.Content()
-	if err != nil {
-		return errors.Wrap(err, "get content of response cache configuration")
-	}
-
-	var cacheConfig *queryrange.ResultsCacheConfig
-	if len(respCacheContentYaml) > 0 {
-		cacheConfig, err = cache.NewResponseCacheConfig(respCacheContentYaml)
-		if err != nil {
-			return errors.Wrap(err, "create response cache")
-		}
-	}
-
-	codec := queryfrontend.NewThanosCodec(conf.queryRangeConfig.partialResponseStrategy)
-	tripperWare, err := queryfrontend.NewTripperWare(
-		limits,
-		cacheConfig,
-		codec,
-		queryrange.PrometheusResponseExtractor{},
-		time.Duration(conf.queryRangeConfig.splitInterval),
-		conf.queryRangeConfig.maxRetries,
-		reg,
-		logger,
-	)
+	tripperWare, err := queryfrontend.NewTripperware(cfg.Config, reg, logger)
 	if err != nil {
 		return errors.Wrap(err, "setup query range middlewares")
 	}
@@ -173,16 +134,16 @@ func runQueryFrontend(
 
 	// Configure Request Logging for HTTP calls.
 	opts := []logging.Option{logging.WithDecider(func() logging.Decision {
-		return logging.LogDecision[conf.requestLoggingDecision]
+		return logging.LogDecision[cfg.RequestLoggingDecision]
 	})}
 	logMiddleware := logging.NewHTTPServerMiddleware(logger, opts...)
 	ins := extpromhttp.NewInstrumentationMiddleware(reg)
 
 	// Start metrics HTTP server.
 	{
 		srv := httpserver.New(logger, reg, comp, httpProbe,
-			httpserver.WithListen(conf.http.bindAddress),
-			httpserver.WithGracePeriod(time.Duration(conf.http.gracePeriod)),
+			httpserver.WithListen(cfg.http.bindAddress),
+			httpserver.WithGracePeriod(time.Duration(cfg.http.gracePeriod)),
 		)
 
 		instr := func(f http.HandlerFunc) http.HandlerFunc {

docs/components/query-frontend.md

@@ -29,7 +29,7 @@ For more information please check out [initial design proposal](https://thanos.i
 ### Splitting
 
 Query Frontend splits a long query into multiple short queries based on the configured `--query-range.split-interval` flag. The default value of `--query-range.split-interval`
-is `24h`. Set it to `0` disables query splitting, but please note that caching is also disabled in this case.
+is `24h`. When caching is enabled it should be greater than `0`.
 
 There are some benefits from query splitting:
 
@@ -44,9 +44,11 @@ Query Frontend supports a retry mechanism to retry query when HTTP requests are
 ### Caching
 
 Query Frontend supports caching query results and reuses them on subsequent queries. If the cached results are incomplete,
-Query Frontend calculates the required subqueries and executes them in parallel on downstream queriers. Query Frontend can optionally align queries with their step parameter to improve the cacheability of the query results.
+Query Frontend calculates the required subqueries and executes them in parallel on downstream queriers.
+Query Frontend can optionally align queries with their step parameter to improve the cacheability of the query results.
+Currently, in-memory cache (fifo cache) and memcached are supported.
 
-Currently, only in-memory cache (fifo cache) is supported. An example config:
+#### In-memory
 
 [embedmd]:# (../flags/config_response_cache_in_memory.txt yaml)
 ```yaml
@@ -57,9 +59,27 @@ config:
   validity: 0s
 ```
 
+#### Memcached
+
+[embedmd]:# (../flags/config_response_cache_memcached.txt yaml)
+```yaml
+type: MEMCACHED
+config:
+  addresses: []
+  timeout: 0s
+  max_idle_connections: 0
+  max_async_concurrency: 0
+  max_async_buffer_size: 0
+  max_get_multi_concurrency: 0
+  max_item_size: 0
+  max_get_multi_batch_size: 0
+  dns_provider_update_interval: 0s
+  expiration: 0s
+```
+
 ### Slow Query Log
 
-Query Frontend supports `--query-frontend.log_queries_longer_than` flag to log queries running longer some duration.
+Query Frontend supports `--query-frontend.log-queries-longer-than` flag to log queries running longer than some duration.
 
 ## Naming
 
@@ -89,9 +109,14 @@ Flags:
                               priority). Content of YAML file with tracing
                               configuration. See format details:
                               https://thanos.io/tip/thanos/tracing.md/#configuration
+      --http-address="0.0.0.0:10902"
+                              Listen host:port for HTTP endpoints.
+      --http-grace-period=2m  Time to wait after an interrupt received for HTTP
+                              Server.
       --query-range.split-interval=24h
                               Split queries by an interval and execute in
-                              parallel, 0 disables it.
+                              parallel, it should be greater than 0 when
+                              response-cache-config is configured.
       --query-range.max-retries-per-request=5
                               Maximum number of retries for a single request;
                               beyond this, the downstream error is returned.
@@ -100,7 +125,7 @@ Flags:
                               the query-frontend, 0 disables it.
       --query-range.max-query-parallelism=14
                               Maximum number of queries will be scheduled in
-                              parallel by the frontend.
+                              parallel by the Frontend.
       --query-range.response-cache-max-freshness=1m
                               Most recent allowed cacheable result, to prevent
                               caching very recent results that might still be in
@@ -117,15 +142,11 @@ Flags:
                               'query-range.response-cache-config-file' flag
                               (lower priority). Content of YAML file that
                               contains response cache configuration.
-      --http-address="0.0.0.0:10902"
-                              Listen host:port for HTTP endpoints.
-      --http-grace-period=2m  Time to wait after an interrupt received for HTTP
-                              Server.
       --query-frontend.downstream-url="http://localhost:9090"
                               URL of downstream Prometheus Query compatible API.
       --query-frontend.compress-responses
                               Compress HTTP responses.
-      --query-frontend.log_queries_longer_than=0
+      --query-frontend.log-queries-longer-than=0
                               Log queries that are slower than the specified
                               duration. Set to 0 to disable. Set to < 0 to
                               enable on all queries.