feat: show health warnings

mtojek · mtojek · commit 6f4c14d1d358 · 2023-11-10T12:08:50.000+01:00
diff --git a/coderd/healthcheck/derphealth/derp.go b/coderd/healthcheck/derphealth/derp.go
@@ -27,6 +27,7 @@ import (
 // @typescript-generate Report
 type Report struct {
 	Healthy bool `json:"healthy"`
+	Warning bool `json:"warning"`
 
 	Regions map[int]*RegionReport `json:"regions"`
 
@@ -41,6 +42,7 @@ type Report struct {
 type RegionReport struct {
 	mu      sync.Mutex
 	Healthy bool `json:"healthy"`
+	Warning bool `json:"warning"`
 
 	Region      *tailcfg.DERPRegion `json:"region"`
 	NodeReports []*NodeReport       `json:"node_reports"`
@@ -53,6 +55,7 @@ type NodeReport struct {
 	clientCounter int
 
 	Healthy bool              `json:"healthy"`
+	Warning bool              `json:"warning"`
 	Node    *tailcfg.DERPNode `json:"node"`
 
 	ServerInfo          derp.ServerInfoMessage `json:"node_info"`
@@ -108,6 +111,9 @@ func (r *Report) Run(ctx context.Context, opts *ReportOptions) {
 			if !regionReport.Healthy {
 				r.Healthy = false
 			}
+			if regionReport.Warning {
+				r.Warning = true
+			}
 			mu.Unlock()
 		}()
 	}
@@ -159,6 +165,9 @@ func (r *RegionReport) Run(ctx context.Context) {
 			if !nodeReport.Healthy {
 				r.Healthy = false
 			}
+			if nodeReport.Warning {
+				r.Warning = true
+			}
 			r.mu.Unlock()
 		}()
 	}
@@ -208,14 +217,15 @@ func (r *NodeReport) Run(ctx context.Context) {
 
 	// We can't exchange messages with the node,
 	if (!r.CanExchangeMessages && !r.Node.STUNOnly) ||
-		// A node may use websockets because `Upgrade: DERP` may be blocked on
-		// the load balancer. This is unhealthy because websockets are slower
-		// than the regular DERP protocol.
-		r.UsesWebsocket ||
 		// The node was marked as STUN compatible but the STUN test failed.
 		r.STUN.Error != nil {
 		r.Healthy = false
 	}
+
+	// A node may use websockets because `Upgrade: DERP` may be blocked on
+	// the load balancer. This is unhealthy because websockets are slower
+	// than the regular DERP protocol.
+	r.Warning = r.UsesWebsocket
 }
 
 func (r *NodeReport) doExchangeMessage(ctx context.Context) {
diff --git a/coderd/healthcheck/healthcheck.go b/coderd/healthcheck/healthcheck.go
@@ -36,6 +36,8 @@ type Report struct {
 	Time time.Time `json:"time"`
 	// Healthy is true if the report returns no errors.
 	Healthy bool `json:"healthy"`
+	// Warning is true when Coder is operational but its performance might be impacted.
+	Warning bool `json:"warning"`
 	// FailingSections is a list of sections that have failed their healthcheck.
 	FailingSections []string `json:"failing_sections"`
 
@@ -156,6 +158,9 @@ func Run(ctx context.Context, opts *ReportOptions) *Report {
 	if !report.DERP.Healthy {
 		report.FailingSections = append(report.FailingSections, SectionDERP)
 	}
+	if report.DERP.Warning {
+		report.Warning = true
+	}
 	if !report.AccessURL.Healthy {
 		report.FailingSections = append(report.FailingSections, SectionAccessURL)
 	}
diff --git a/coderd/healthcheck/healthcheck_test.go b/coderd/healthcheck/healthcheck_test.go
@@ -40,6 +40,7 @@ func TestHealthcheck(t *testing.T) {
 		name            string
 		checker         *testChecker
 		healthy         bool
+		warning         bool
 		failingSections []string
 	}{{
 		name: "OK",
@@ -77,6 +78,26 @@ func TestHealthcheck(t *testing.T) {
 		},
 		healthy:         false,
 		failingSections: []string{healthcheck.SectionDERP},
+	}, {
+		name: "DERPWarning",
+		checker: &testChecker{
+			DERPReport: derphealth.Report{
+				Healthy: true,
+				Warning: true,
+			},
+			AccessURLReport: healthcheck.AccessURLReport{
+				Healthy: true,
+			},
+			WebsocketReport: healthcheck.WebsocketReport{
+				Healthy: true,
+			},
+			DatabaseReport: healthcheck.DatabaseReport{
+				Healthy: true,
+			},
+		},
+		healthy:         true,
+		warning:         true,
+		failingSections: nil,
 	}, {
 		name: "AccessURLFail",
 		checker: &testChecker{
@@ -151,8 +172,10 @@ func TestHealthcheck(t *testing.T) {
 			})
 
 			assert.Equal(t, c.healthy, report.Healthy)
+			assert.Equal(t, c.warning, report.Warning)
 			assert.Equal(t, c.failingSections, report.FailingSections)
 			assert.Equal(t, c.checker.DERPReport.Healthy, report.DERP.Healthy)
+			assert.Equal(t, c.checker.DERPReport.Warning, report.DERP.Warning)
 			assert.Equal(t, c.checker.AccessURLReport.Healthy, report.AccessURL.Healthy)
 			assert.Equal(t, c.checker.WebsocketReport.Healthy, report.Websocket.Healthy)
 			assert.NotZero(t, report.Time)
