Merge pull request #55 from subnova-etsy/clearfrequencysec

Add support for specifying the ClearFrequencySec attribute

README.md

@@ -65,7 +65,7 @@ The redis host should be a hostname and a port, for example `redis.mydomain.com:
 
 ## How sampling decisions are made
 
-In the configuration file, there is a place to choose a sampling method and some options for each. The `DynamicSampler` is the most interesting and most commonly used, so that's the one that gets described here. It uses the `AvgSampleRate` algorithm from the [`dynsampler-go`](https://github.com/honeycombio/dynsampler-go) package. Briefly described, you configure samproxy to examine the trace for a set of fields (for example, `request.status_code` and `request.method`). It collects all the values found in those fields anywhere in the trace (eg "200" and "GET") together in to a key it hands to the dynsampler. The dynsampler code will look at the frequency that key appears during the previous 30 seconds and use that to hand back a desired sample rate. More frequent keys are sampled more heavily, so that an even distribution of traffic across the keyspace is represented in Honeycomb.
+In the configuration file, there is a place to choose a sampling method and some options for each. The `DynamicSampler` is the most interesting and most commonly used, so that's the one that gets described here. It uses the `AvgSampleRate` algorithm from the [`dynsampler-go`](https://github.com/honeycombio/dynsampler-go) package. Briefly described, you configure samproxy to examine the trace for a set of fields (for example, `request.status_code` and `request.method`). It collects all the values found in those fields anywhere in the trace (eg "200" and "GET") together in to a key it hands to the dynsampler. The dynsampler code will look at the frequency that key appears during the previous 30 seconds (or other value set by the `ClearFrequencySec` setting) and use that to hand back a desired sample rate. More frequent keys are sampled more heavily, so that an even distribution of traffic across the keyspace is represented in Honeycomb.
 
 By selecting fields well, you can drop significant amounts of traffic while still retaining good visibility into the areas of traffic that interest you. For example, if you want to make sure you have a complete list of all URL handlers invoked, you would add the URL (or a normalized form) as one of the fields to include. Be careful in your selection though, because if the combination of fields cretes a unique key each time, you won't sample out any traffic. Because of this it is not effective to use fields that have unique values (like a UUID) as one of the sampling fields. Each field included should ideally have values that appear many times within any given 30 second window in order to effectively turn in to a sample rate.
 

config.toml

@@ -169,7 +169,7 @@ CacheCapacity = 1000
 		# implementation. This sampler collects the values of a number of fields from a
 		# trace and uses them to form a key. This key is handed to the standard dynamic
 		# sampler algorithm which generates a sample rate based on the frequency with
-		# which that key has appeared in the previous 30 seconds. See
+		# which that key has appeared in the previous ClearFrequencySec seconds. See
 		# https://github.com/honeycombio/dynsampler-go for more detail on the mechanics
 		# of the dynamic sampler.  This sampler uses the AvgSampleRate algorithm from
 		# that package.
@@ -219,6 +219,12 @@ CacheCapacity = 1000
 		# AddSampleRateKeyToTrace is true.
 		AddSampleRateKeyToTraceField = "meta.samproxy.dynsampler_key"
 
+        # ClearFrequencySec is the name of the field the sampler will use to determine
+        # the period over which it will calculate the sample rate. This setting defaults
+        # to 30.
+        # Eligible for live reload.
+		ClearFrequencySec = 60
+
 	[SamplerConfig.dataset2]
 
 		Sampler = "DeterministicSampler"

config/config_test.go

@@ -29,6 +29,7 @@ func TestGetSamplerTypes(t *testing.T) {
 		UseTraceLength = true
 		AddSampleRateKeyToTrace = true
 		AddSampleRateKeyToTraceField = "meta.samproxy.dynsampler_key"
+		ClearFrequencySec = 60
 
 	[SamplerConfig.dataset2]
 

sample/dynamic.go

@@ -20,6 +20,7 @@ type DynamicSampler struct {
 	Metrics metrics.Metrics
 
 	sampleRate        int64
+	clearFrequencySec int64
 	fieldList         []string
 	useTraceLength    bool
 	addDynsampleKey   bool
@@ -31,6 +32,7 @@ type DynamicSampler struct {
 
 type DynSamplerConfig struct {
 	SampleRate                   int64
+	ClearFrequencySec			 int64
 	FieldList                    []string
 	UseTraceLength               bool
 	AddSampleRateKeyToTrace      bool
@@ -51,6 +53,10 @@ func (d *DynamicSampler) Start() error {
 		dsConfig.SampleRate = 1
 	}
 	d.sampleRate = dsConfig.SampleRate
+	if dsConfig.ClearFrequencySec == 0 {
+		dsConfig.ClearFrequencySec = 30
+	}
+	d.clearFrequencySec = dsConfig.ClearFrequencySec
 
 	// get list of fields to use when constructing the dynsampler key
 	fieldList := dsConfig.FieldList
@@ -67,6 +73,7 @@ func (d *DynamicSampler) Start() error {
 	// spin up the actual dynamic sampler
 	d.dynsampler = &dynsampler.AvgSampleRate{
 		GoalSampleRate: int(d.sampleRate),
+		ClearFrequencySec: int(d.clearFrequencySec),
 	}
 	d.dynsampler.Start()
 
@@ -99,6 +106,13 @@ func (d *DynamicSampler) reloadConfigs() {
 		configChanged = true
 		d.sampleRate = dsConfig.SampleRate
 	}
+	if dsConfig.ClearFrequencySec == 0 {
+		dsConfig.ClearFrequencySec = 30
+	}
+	if d.clearFrequencySec != dsConfig.ClearFrequencySec {
+		configChanged = true
+		d.clearFrequencySec = dsConfig.ClearFrequencySec
+	}
 
 	// get list of fields to use when constructing the dynsampler key
 	fieldList := dsConfig.FieldList
@@ -135,6 +149,7 @@ func (d *DynamicSampler) reloadConfigs() {
 	if configChanged {
 		newSampler := &dynsampler.AvgSampleRate{
 			GoalSampleRate: int(d.sampleRate),
+			ClearFrequencySec: int(d.clearFrequencySec),
 		}
 		newSampler.Start()