Move to json storage, address pblindex issue (#721)

This patch changes the way we persist the data on disk when
running Calico/VPP. Instead of using struc and binary format
we transition to json files. Size should not be an issue as number
of pods per node are typically low (~100). This will make
troubleshooting easier and errors clearer when parsing fails.

We thus remove the /bin/debug troubleshooting utility as the
data format is not human readable.

Doing this, we address an issue where PBL indexes were reused
upon dataplane restart, as they were stored in a list. We now
will use a map to retain the containerIP mapping.

We also split the configuration from runtime spec in LocalPodSpec
and add a step to clear it when corresponding VRFs are not found
in VPP.

Finally we address an issue where uRPF was not properly set up
for ipv6.

Signed-off-by: Nathan Skrzypczak <[email protected]>

Author: sknat

calico-vpp-agent/Dockerfile

@@ -3,7 +3,6 @@ FROM ubuntu:22.04
 LABEL maintainer="[email protected]"
 
 ADD bin/gobgp /bin/gobgp
-ADD bin/debug /bin/debug
 ADD version /etc/calicovppversion
 ADD bin/felix-api-proxy /bin/felix-api-proxy
 ADD bin/calico-vpp-agent /bin/calico-vpp-agent

calico-vpp-agent/Makefile

@@ -18,7 +18,6 @@ felix-api-proxy: bin
 
 build: felix-api-proxy bin
 	${DOCKER_RUN} go build -o ./bin/calico-vpp-agent ./cmd
-	${DOCKER_RUN} go build -o ./bin/debug ./cmd/debug-state
 
 gobgp: bin
 	${DOCKER_RUN} go build -o ./bin/gobgp github.com/osrg/gobgp/v3/cmd/gobgp/

calico-vpp-agent/cmd/calico_vpp_dataplane.go

@@ -19,6 +19,7 @@ import (
 	"context"
 	"os"
 	"os/signal"
+	"runtime/coverage"
 	"syscall"
 	"time"
 
@@ -219,20 +220,41 @@ func main() {
 
 	log.Infof("Agent started")
 
-	interruptSignalChannel := make(chan os.Signal, 2)
-	signal.Notify(interruptSignalChannel, os.Interrupt, syscall.SIGTERM)
-
-	usr1SignalChannel := make(chan os.Signal, 2)
-	signal.Notify(usr1SignalChannel, syscall.SIGUSR1)
+	sigChan := make(chan os.Signal, 2)
+	signal.Notify(sigChan,
+		os.Interrupt,
+		syscall.SIGTERM,
+		syscall.SIGUSR1,
+		syscall.SIGUSR2,
+	)
 
 	select {
-	case <-usr1SignalChannel:
-		/* vpp-manager pokes us with USR1 if VPP terminates */
-		log.Warnf("Vpp stopped, exiting...")
-		t.Kill(errors.Errorf("Caught signal USR1"))
-	case <-interruptSignalChannel:
-		log.Infof("SIG received, exiting")
-		t.Kill(errors.Errorf("Caught INT signal"))
+	case sig := <-sigChan:
+		switch sig {
+		case os.Interrupt:
+			fallthrough
+		case syscall.SIGTERM:
+			log.Infof("SIG received, exiting")
+			t.Kill(errors.Errorf("Caught INT signal"))
+		case syscall.SIGUSR1:
+			// vpp-manager pokes us with USR1 if VPP terminates
+			log.Warnf("Vpp stopped, exiting...")
+			t.Kill(errors.Errorf("Caught signal USR1"))
+		case syscall.SIGUSR2:
+			// the USR2 signal outputs the coverage data,
+			// provided the binary is compiled with -cover and
+			// GOCOVERDIR is set. This allows us to not require
+			// a proper binary termination in order to get coverage data.
+			log.Warn("Received SIGUSR2, writing coverage")
+			err := coverage.WriteCountersDir(os.Getenv("GOCOVERDIR"))
+			if err != nil {
+				log.WithError(err).Error("Could not write counters dir")
+			}
+			err = coverage.WriteMetaDir(os.Getenv("GOCOVERDIR"))
+			if err != nil {
+				log.WithError(err).Error("Could not write meta dir")
+			}
+		}
 	case <-t.Dying():
 		log.Errorf("tomb Dying %s", t.Err())
 	}

calico-vpp-agent/cmd/debug-state/debug-state.go

@@ -182,7 +182,7 @@ var _ = Describe("Pod-related functionality of CNI", func() {
 						Workload: &cniproto.WorkloadIDs{
 							Annotations: map[string]string{
 								// needed just for setting up steering of traffic to default Tun/Tap and to secondary Memif
-								cni.VppAnnotationPrefix + cni.MemifPortAnnotation: fmt.Sprintf("tcp:%d-%d,udp:%d-%d",
+								config.MemifPortAnnotation: fmt.Sprintf("tcp:%d-%d,udp:%d-%d",
 									memifTCPPortStart, memifTCPPortEnd, memifUDPPortStart, memifUDPPortEnd),
 							},
 						},
@@ -418,7 +418,7 @@ var _ = Describe("Pod-related functionality of CNI", func() {
 							Workload: &cniproto.WorkloadIDs{
 								Annotations: map[string]string{
 									// needed just for setting up steering of traffic to default Tun/Tap and to secondary Memif
-									cni.VppAnnotationPrefix + cni.MemifPortAnnotation: fmt.Sprintf("tcp:%d-%d,udp:%d-%d",
+									config.MemifPortAnnotation: fmt.Sprintf("tcp:%d-%d,udp:%d-%d",
 										memifTCPPortStart, memifTCPPortEnd, memifUDPPortStart, memifUDPPortEnd),
 								},
 							},