feat(ooniprobe): introduce websites_max_runtime (#273)

We cannot control anymore the maximum number of URLs using the API
because now we are using check-in, that has no such limit.

We could theoretically clamp the number of URLs to measure after
the call to check-in, and still honour the setting.

Yet, the right thing to do seems to introduce a max runtime variable
because that is what desktop and mobile do.

Thus, introduce code that warns the user about the change in the
settings, should they have set the URL limit to nonzero.

We are going to do a best effort conversion from the URL limit to
the maximum runtime for the rest of 2021.

Since then, we will silently ignore the URL limit.

This work is part of https://github.com/ooni/probe/issues/1299.

cmd/ooniprobe/internal/config/settings.go

@@ -1,38 +1,5 @@
 package config
 
-var websiteCategories = []string{
-	"ALDR",
-	"ANON",
-	"COMM",
-	"COMT",
-	"CTRL",
-	"CULTR",
-	"DATE",
-	"ECON",
-	"ENV",
-	"FILE",
-	"GAME",
-	"GMB",
-	"GOVT",
-	"GRP",
-	"HACK",
-	"HATE",
-	"HOST",
-	"HUMR",
-	"IGO",
-	"LGBT",
-	"MILX",
-	"MMED",
-	"NEWS",
-	"POLR",
-	"PORN",
-	"PROV",
-	"PUBH",
-	"REL",
-	"SRCH",
-	"XED",
-}
-
 // Sharing settings
 type Sharing struct {
 	UploadResults bool `json:"upload_results"`
@@ -45,6 +12,7 @@ type Advanced struct {
 
 // Nettests related settings
 type Nettests struct {
+	WebsitesMaxRuntime           int64    `json:"websites_max_runtime"`
 	WebsitesURLLimit             int64    `json:"websites_url_limit"`
 	WebsitesEnabledCategoryCodes []string `json:"websites_enabled_category_codes"`
 }

cmd/ooniprobe/internal/config/testdata/valid-config.json

@@ -5,7 +5,7 @@
     "upload_results": true
   },
   "nettests": {
-    "websites_url_limit": 0
+    "websites_max_runtime": 0
   },
   "advanced": {
     "send_crash_reports": true

cmd/ooniprobe/internal/nettests/nettests.go

@@ -111,10 +111,16 @@ func (c *Controller) Run(builder *engine.ExperimentBuilder, inputs []string) err
 		}
 	}
 
-	c.ntStartTime = time.Now()
+	maxRuntime := time.Duration(c.Probe.Config().Nettests.WebsitesMaxRuntime) * time.Second
+	start := time.Now()
+	c.ntStartTime = start
 	for idx, input := range inputs {
-		if c.Probe.IsTerminated() == true {
-			log.Debug("isTerminated == true, breaking the input loop")
+		if c.Probe.IsTerminated() {
+			log.Info("user requested us to terminate using Ctrl-C")
+			break
+		}
+		if maxRuntime > 0 && time.Since(start) > maxRuntime {
+			log.Info("exceeded maximum runtime")
 			break
 		}
 		c.curInputIdx = idx // allow for precise progress

cmd/ooniprobe/internal/nettests/run.go

@@ -1,6 +1,8 @@
 package nettests
 
 import (
+	"time"
+
 	"github.com/apex/log"
 	"github.com/ooni/probe-cli/v3/cmd/ooniprobe/internal/database"
 	"github.com/ooni/probe-cli/v3/cmd/ooniprobe/internal/ooni"
@@ -15,9 +17,36 @@ type RunGroupConfig struct {
 	Inputs     []string
 }
 
+const websitesURLLimitRemoved = `WARNING: CONFIGURATION CHANGE REQUIRED:
+
+* Since ooniprobe 3.9.0, websites_url_limit has been replaced
+  by websites_max_runtime in the configuration
+
+* To silence this warning either set websites_url_limit to zero or
+  replace it with websites_max_runtime
+
+* For the rest of 2021, we will automatically convert websites_url_limit
+  to websites_max_runtime (if the latter is not already set)
+
+* We will consider that each URL in websites_url_limit takes five
+  seconds to run and thus calculate websites_max_runtime
+
+* Since 2022, we will start silently ignoring websites_url_limit
+`
+
 // RunGroup runs a group of nettests according to the specified config.
 func RunGroup(config RunGroupConfig) error {
-	if config.Probe.IsTerminated() == true {
+	if config.Probe.Config().Nettests.WebsitesURLLimit > 0 {
+		log.Warn(websitesURLLimitRemoved)
+		if config.Probe.Config().Nettests.WebsitesMaxRuntime <= 0 {
+			limit := config.Probe.Config().Nettests.WebsitesURLLimit
+			maxRuntime := 5 * limit
+			config.Probe.Config().Nettests.WebsitesMaxRuntime = maxRuntime
+		}
+		time.Sleep(30 * time.Second)
+	}
+
+	if config.Probe.IsTerminated() {
 		log.Debugf("context is terminated, stopping runNettestGroup early")
 		return nil
 	}

cmd/ooniprobe/internal/nettests/web_connectivity.go

@@ -9,16 +9,10 @@ import (
 	"github.com/ooni/probe-cli/v3/internal/engine/model"
 )
 
-// TODO(bassosimone): we should remove the limit argument and
-// we should also remove it from the config.
-
 // TODO(bassosimone): we should propagate the kind of run
 // to here such that we get the right runType.
 
-// TODO(bassosimone): we are breaking the use case in which
-// someone choose the number of URLs explicitly via the config.
-
-func lookupURLs(ctl *Controller, limit int64, categories []string) ([]string, map[int64]int64, error) {
+func lookupURLs(ctl *Controller, categories []string) ([]string, map[int64]int64, error) {
 	inputloader := &engine.InputLoader{
 		CheckInConfig: &model.CheckInConfig{
 			WebConnectivity: model.CheckInConfigWebConnectivity{
@@ -53,13 +47,12 @@ func lookupURLs(ctl *Controller, limit int64, categories []string) ([]string, ma
 }
 
 // WebConnectivity test implementation
-type WebConnectivity struct {
-}
+type WebConnectivity struct{}
 
 // Run starts the test
 func (n WebConnectivity) Run(ctl *Controller) error {
 	log.Debugf("Enabled category codes are the following %v", ctl.Probe.Config().Nettests.WebsitesEnabledCategoryCodes)
-	urls, urlIDMap, err := lookupURLs(ctl, ctl.Probe.Config().Nettests.WebsitesURLLimit, ctl.Probe.Config().Nettests.WebsitesEnabledCategoryCodes)
+	urls, urlIDMap, err := lookupURLs(ctl, ctl.Probe.Config().Nettests.WebsitesEnabledCategoryCodes)
 	if err != nil {
 		return err
 	}

cmd/ooniprobe/internal/ooni/default-config.json

@@ -5,7 +5,7 @@
     "upload_results": true
   },
   "nettests": {
-    "websites_url_limit": 0
+    "websites_max_runtime": 0
   },
   "advanced": {
     "send_crash_reports": true

cmd/ooniprobe/testdata/testing-config.json

@@ -5,7 +5,7 @@
     "upload_results": true
   },
   "nettests": {
-    "websites_url_limit": 10
+    "websites_max_runtime": 15
   },
   "advanced": {
     "send_crash_reports": true

debian/ooniprobe.conf.disabled

@@ -6,7 +6,7 @@
     "upload_results": true
   },
   "nettests": {
-    "websites_url_limit": 0,
+    "websites_max_runtime": 0,
     "websites_enabled_category_codes": null
   },
   "advanced": {