Debugging Envoy and Istiod

Istio provides two very valuable commands to help diagnose traffic management configuration problems, the proxy-status and proxy-config commands. The proxy-status command allows you to get an overview of your mesh and identify the proxy causing the problem. Then proxy-config can be used to inspect Envoy configuration and diagnose the issue.

If you want to try the commands described below, you can either:

OR

  • Use similar commands against your own application running in a Kubernetes cluster.

Get an overview of your mesh

The proxy-status command allows you to get an overview of your mesh. If you suspect one of your sidecars isn’t receiving configuration or is out of sync then proxy-status will tell you this.

  1. $ istioctl proxy-status
  2. NAME CDS LDS EDS RDS ISTIOD VERSION
  3. details-v1-558b8b4b76-qzqsg.default SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0
  4. istio-ingressgateway-66c994c45c-cmb7x.istio-system SYNCED SYNCED SYNCED NOT SENT istiod-6cf8d4f9cb-wm7x6 1.7.0
  5. productpage-v1-6987489c74-nc7tj.default SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0
  6. prometheus-7bdc59c94d-hcp59.istio-system SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0
  7. ratings-v1-7dc98c7588-5m6xj.default SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0
  8. reviews-v1-7f99cc4496-rtsqn.default SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0
  9. reviews-v2-7d79d5bd5d-tj6kf.default SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0
  10. reviews-v3-7dbcdcbc56-t8wrx.default SYNCED SYNCED SYNCED SYNCED istiod-6cf8d4f9cb-wm7x6 1.7.0

If a proxy is missing from this list it means that it is not currently connected to a Istiod instance so will not be receiving any configuration.

  • SYNCED means that Envoy has acknowledged the last configuration Istiod has sent to it.
  • NOT SENT means that Istiod hasn’t sent anything to Envoy. This usually is because Istiod has nothing to send.
  • STALE means that Istiod has sent an update to Envoy but has not received an acknowledgement. This usually indicates a networking issue between Envoy and Istiod or a bug with Istio itself.

Retrieve diffs between Envoy and Istiod

The proxy-status command can also be used to retrieve a diff between the configuration Envoy has loaded and the configuration Istiod would send, by providing a proxy ID. This can help you determine exactly what is out of sync and where the issue may lie.

  1. $ istioctl proxy-status details-v1-6dcc6fbb9d-wsjz4.default
  2. --- Istiod Clusters
  3. +++ Envoy Clusters
  4. @@ -374,36 +374,14 @@
  5. "edsClusterConfig": {
  6. "edsConfig": {
  7. "ads": {
  8. }
  9. },
  10. "serviceName": "outbound|443||public-cr0bdc785ce3f14722918080a97e1f26be-alb1.kube-system.svc.cluster.local"
  11. - },
  12. - "connectTimeout": "1.000s",
  13. - "circuitBreakers": {
  14. - "thresholds": [
  15. - {
  16. -
  17. - }
  18. - ]
  19. - }
  20. - }
  21. - },
  22. - {
  23. - "cluster": {
  24. - "name": "outbound|53||kube-dns.kube-system.svc.cluster.local",
  25. - "type": "EDS",
  26. - "edsClusterConfig": {
  27. - "edsConfig": {
  28. - "ads": {
  29. -
  30. - }
  31. - },
  32. - "serviceName": "outbound|53||kube-dns.kube-system.svc.cluster.local"
  33. },
  34. "connectTimeout": "1.000s",
  35. "circuitBreakers": {
  36. "thresholds": [
  37. {
  38. }
  39. Listeners Match
  40. Routes Match (RDS last loaded at Tue, 04 Aug 2020 11:52:54 IST)

Here you can see that the listeners and routes match but the clusters are out of sync.

Deep dive into Envoy configuration

The proxy-config command can be used to see how a given Envoy instance is configured. This can then be used to pinpoint any issues you are unable to detect by just looking through your Istio configuration and custom resources. To get a basic summary of clusters, listeners or routes for a given pod use the command as follows (changing clusters for listeners or routes when required):

  1. $ istioctl proxy-config cluster -n istio-system istio-ingressgateway-7d6874b48f-qxhn5
  2. SERVICE FQDN PORT SUBSET DIRECTION TYPE DESTINATION RULE
  3. BlackHoleCluster - - - STATIC
  4. agent - - - STATIC
  5. details.default.svc.cluster.local 9080 - outbound EDS details.default
  6. istio-ingressgateway.istio-system.svc.cluster.local 80 - outbound EDS
  7. istio-ingressgateway.istio-system.svc.cluster.local 443 - outbound EDS
  8. istio-ingressgateway.istio-system.svc.cluster.local 15021 - outbound EDS
  9. istio-ingressgateway.istio-system.svc.cluster.local 15443 - outbound EDS
  10. istiod.istio-system.svc.cluster.local 443 - outbound EDS
  11. istiod.istio-system.svc.cluster.local 853 - outbound EDS
  12. istiod.istio-system.svc.cluster.local 15010 - outbound EDS
  13. istiod.istio-system.svc.cluster.local 15012 - outbound EDS
  14. istiod.istio-system.svc.cluster.local 15014 - outbound EDS
  15. kube-dns.kube-system.svc.cluster.local 53 - outbound EDS
  16. kube-dns.kube-system.svc.cluster.local 9153 - outbound EDS
  17. kubernetes.default.svc.cluster.local 443 - outbound EDS
  18. ...
  19. productpage.default.svc.cluster.local 9080 - outbound EDS
  20. prometheus_stats - - - STATIC
  21. ratings.default.svc.cluster.local 9080 - outbound EDS
  22. reviews.default.svc.cluster.local 9080 - outbound EDS
  23. sds-grpc - - - STATIC
  24. xds-grpc - - - STRICT_DNS
  25. zipkin - - - STRICT_DNS

In order to debug Envoy you need to understand Envoy clusters/listeners/routes/endpoints and how they all interact. We will use the proxy-config command with the -o json and filtering flags to follow Envoy as it determines where to send a request from the productpage pod to the reviews pod at reviews:9080.

  1. If you query the listener summary on a pod you will notice Istio generates the following listeners:

    • A listener on 0.0.0.0:15006 that receives all inbound traffic to the pod and a listener on 0.0.0.0:15001 that receives all outbound traffic to the pod, then hands the request over to a virtual listener.
    • A virtual listener per service IP, per each non-HTTP for outbound TCP/HTTPS traffic.
    • A virtual listener on the pod IP for each exposed port for inbound traffic.
    • A virtual listener on 0.0.0.0 per each HTTP port for outbound HTTP traffic.
    1. $ istioctl proxy-config listeners productpage-v1-6c886ff494-7vxhs
    2. ADDRESS PORT MATCH DESTINATION
    3. 10.96.0.10 53 ALL Cluster: outbound|53||kube-dns.kube-system.svc.cluster.local
    4. 0.0.0.0 80 App: HTTP Route: 80
    5. 0.0.0.0 80 ALL PassthroughCluster
    6. 10.100.93.102 443 ALL Cluster: outbound|443||istiod.istio-system.svc.cluster.local
    7. 10.111.121.13 443 ALL Cluster: outbound|443||istio-ingressgateway.istio-system.svc.cluster.local
    8. 10.96.0.1 443 ALL Cluster: outbound|443||kubernetes.default.svc.cluster.local
    9. 10.100.93.102 853 App: HTTP Route: istiod.istio-system.svc.cluster.local:853
    10. 10.100.93.102 853 ALL Cluster: outbound|853||istiod.istio-system.svc.cluster.local
    11. 0.0.0.0 9080 App: HTTP Route: 9080
    12. 0.0.0.0 9080 ALL PassthroughCluster
    13. 0.0.0.0 9090 App: HTTP Route: 9090
    14. 0.0.0.0 9090 ALL PassthroughCluster
    15. 10.96.0.10 9153 App: HTTP Route: kube-dns.kube-system.svc.cluster.local:9153
    16. 10.96.0.10 9153 ALL Cluster: outbound|9153||kube-dns.kube-system.svc.cluster.local
    17. 0.0.0.0 15001 ALL PassthroughCluster
    18. 0.0.0.0 15006 Addr: 10.244.0.22/32:15021 inbound|15021|mgmt-15021|mgmtCluster
    19. 0.0.0.0 15006 Addr: 10.244.0.22/32:9080 Inline Route: /*
    20. 0.0.0.0 15006 Trans: tls; App: HTTP TLS; Addr: 0.0.0.0/0 Inline Route: /*
    21. 0.0.0.0 15006 App: HTTP; Addr: 0.0.0.0/0 Inline Route: /*
    22. 0.0.0.0 15006 App: Istio HTTP Plain; Addr: 10.244.0.22/32:9080 Inline Route: /*
    23. 0.0.0.0 15006 Addr: 0.0.0.0/0 InboundPassthroughClusterIpv4
    24. 0.0.0.0 15006 Trans: tls; App: TCP TLS; Addr: 0.0.0.0/0 InboundPassthroughClusterIpv4
    25. 0.0.0.0 15010 App: HTTP Route: 15010
    26. 0.0.0.0 15010 ALL PassthroughCluster
    27. 10.100.93.102 15012 ALL Cluster: outbound|15012||istiod.istio-system.svc.cluster.local
    28. 0.0.0.0 15014 App: HTTP Route: 15014
    29. 0.0.0.0 15014 ALL PassthroughCluster
    30. 0.0.0.0 15021 ALL Inline Route: /healthz/ready*
    31. 10.111.121.13 15021 App: HTTP Route: istio-ingressgateway.istio-system.svc.cluster.local:15021
    32. 10.111.121.13 15021 ALL Cluster: outbound|15021||istio-ingressgateway.istio-system.svc.cluster.local
    33. 0.0.0.0 15090 ALL Inline Route: /stats/prometheus*
    34. 10.111.121.13 15443 ALL Cluster: outbound|15443||istio-ingressgateway.istio-system.svc.cluster.local
  2. From the above summary you can see that every sidecar has a listener bound to 0.0.0.0:15006 which is where IP tables routes all inbound pod traffic to and a listener bound to 0.0.0.0:15001 which is where IP tables routes all outbound pod traffic to. The 0.0.0.0:15001 listener hands the request over to the virtual listener that best matches the original destination of the request, if it can find a matching one. Otherwise, it sends the request to the PassthroughCluster which connects to the destination directly.

    1. $ istioctl proxy-config listeners productpage-v1-6c886ff494-7vxhs --port 15001 -o json
    2. [
    3. {
    4. "name": "virtualOutbound",
    5. "address": {
    6. "socketAddress": {
    7. "address": "0.0.0.0",
    8. "portValue": 15001
    9. }
    10. },
    11. "filterChains": [
    12. {
    13. "filters": [
    14. {
    15. "name": "istio.stats",
    16. "typedConfig": {
    17. "@type": "type.googleapis.com/udpa.type.v1.TypedStruct",
    18. "typeUrl": "type.googleapis.com/envoy.extensions.filters.network.wasm.v3.Wasm",
    19. "value": {
    20. "config": {
    21. "configuration": "{\n \"debug\": \"false\",\n \"stat_prefix\": \"istio\"\n}\n",
    22. "root_id": "stats_outbound",
    23. "vm_config": {
    24. "code": {
    25. "local": {
    26. "inline_string": "envoy.wasm.stats"
    27. }
    28. },
    29. "runtime": "envoy.wasm.runtime.null",
    30. "vm_id": "tcp_stats_outbound"
    31. }
    32. }
    33. }
    34. }
    35. },
    36. {
    37. "name": "envoy.tcp_proxy",
    38. "typedConfig": {
    39. "@type": "type.googleapis.com/envoy.config.filter.network.tcp_proxy.v2.TcpProxy",
    40. "statPrefix": "PassthroughCluster",
    41. "cluster": "PassthroughCluster"
    42. }
    43. }
    44. ],
    45. "name": "virtualOutbound-catchall-tcp"
    46. }
    47. ],
    48. "trafficDirection": "OUTBOUND",
    49. "hiddenEnvoyDeprecatedUseOriginalDst": true
    50. }
    51. ]
  3. Our request is an outbound HTTP request to port 9080 this means it gets handed off to the 0.0.0.0:9080 virtual listener. This listener then looks up the route configuration in its configured RDS. In this case it will be looking up route 9080 in RDS configured by Istiod (via ADS).

    1. $ istioctl proxy-config listeners productpage-v1-6c886ff494-7vxhs -o json --address 0.0.0.0 --port 9080
    2. ...
    3. "rds": {
    4. "configSource": {
    5. "ads": {},
    6. "resourceApiVersion": "V3"
    7. },
    8. "routeConfigName": "9080"
    9. }
    10. ...
  4. The 9080 route configuration only has a virtual host for each service. Our request is heading to the reviews service so Envoy will select the virtual host to which our request matches a domain. Once matched on domain Envoy looks for the first route that matches the request. In this case we don’t have any advanced routing so there is only one route that matches on everything. This route tells Envoy to send the request to the outbound|9080||reviews.default.svc.cluster.local cluster.

    1. $ istioctl proxy-config routes productpage-v1-6c886ff494-7vxhs --name 9080 -o json
    2. [
    3. {
    4. "name": "9080",
    5. "virtualHosts": [
    6. {
    7. "name": "reviews.default.svc.cluster.local:9080",
    8. "domains": [
    9. "reviews.default.svc.cluster.local",
    10. "reviews.default.svc.cluster.local:9080",
    11. "reviews",
    12. "reviews:9080",
    13. "reviews.default.svc.cluster",
    14. "reviews.default.svc.cluster:9080",
    15. "reviews.default.svc",
    16. "reviews.default.svc:9080",
    17. "reviews.default",
    18. "reviews.default:9080",
    19. "10.98.88.0",
    20. "10.98.88.0:9080"
    21. ],
    22. "routes": [
    23. {
    24. "name": "default",
    25. "match": {
    26. "prefix": "/"
    27. },
    28. "route": {
    29. "cluster": "outbound|9080||reviews.default.svc.cluster.local",
    30. "timeout": "0s",
    31. }
    32. }
    33. ]
    34. ...
  5. This cluster is configured to retrieve the associated endpoints from Istiod (via ADS). So Envoy will then use the serviceName field as a key to look up the list of Endpoints and proxy the request to one of them.

    1. $ istioctl proxy-config cluster productpage-v1-6c886ff494-7vxhs --fqdn reviews.default.svc.cluster.local -o json
    2. [
    3. {
    4. "name": "outbound|9080||reviews.default.svc.cluster.local",
    5. "type": "EDS",
    6. "edsClusterConfig": {
    7. "edsConfig": {
    8. "ads": {},
    9. "resourceApiVersion": "V3"
    10. },
    11. "serviceName": "outbound|9080||reviews.default.svc.cluster.local"
    12. },
    13. "connectTimeout": "10s",
    14. "circuitBreakers": {
    15. "thresholds": [
    16. {
    17. "maxConnections": 4294967295,
    18. "maxPendingRequests": 4294967295,
    19. "maxRequests": 4294967295,
    20. "maxRetries": 4294967295
    21. }
    22. ]
    23. },
    24. }
    25. ]
  6. To see the endpoints currently available for this cluster use the proxy-config endpoints command.

    1. $ istioctl proxy-config endpoints productpage-v1-6c886ff494-7vxhs --cluster "outbound|9080||reviews.default.svc.cluster.local"
    2. ENDPOINT STATUS OUTLIER CHECK CLUSTER
    3. 172.17.0.7:9080 HEALTHY OK outbound|9080||reviews.default.svc.cluster.local
    4. 172.17.0.8:9080 HEALTHY OK outbound|9080||reviews.default.svc.cluster.local
    5. 172.17.0.9:9080 HEALTHY OK outbound|9080||reviews.default.svc.cluster.local

Inspecting bootstrap configuration

So far we have looked at configuration retrieved (mostly) from Istiod, however Envoy requires some bootstrap configuration that includes information like where Istiod can be found. To view this use the following command:

  1. $ istioctl proxy-config bootstrap -n istio-system istio-ingressgateway-7d6874b48f-qxhn5
  2. {
  3. "bootstrap": {
  4. "node": {
  5. "id": "router~172.30.86.14~istio-ingressgateway-7d6874b48f-qxhn5.istio-system~istio-system.svc.cluster.local",
  6. "cluster": "istio-ingressgateway",
  7. "metadata": {
  8. "CLUSTER_ID": "Kubernetes",
  9. "EXCHANGE_KEYS": "NAME,NAMESPACE,INSTANCE_IPS,LABELS,OWNER,PLATFORM_METADATA,WORKLOAD_NAME,MESH_ID,SERVICE_ACCOUNT,CLUSTER_ID",
  10. "INSTANCE_IPS": "10.244.0.7",
  11. "ISTIO_PROXY_SHA": "istio-proxy:f98b7e538920abc408fbc91c22a3b32bc854d9dc",
  12. "ISTIO_VERSION": "1.7.0",
  13. "LABELS": {
  14. "app": "istio-ingressgateway",
  15. "chart": "gateways",
  16. "heritage": "Tiller",
  17. "istio": "ingressgateway",
  18. "pod-template-hash": "68bf7d7f94",
  19. "release": "istio",
  20. "service.istio.io/canonical-name": "istio-ingressgateway",
  21. "service.istio.io/canonical-revision": "latest"
  22. },
  23. "MESH_ID": "cluster.local",
  24. "NAME": "istio-ingressgateway-68bf7d7f94-sp226",
  25. "NAMESPACE": "istio-system",
  26. "OWNER": "kubernetes://apis/apps/v1/namespaces/istio-system/deployments/istio-ingressgateway",
  27. "ROUTER_MODE": "sni-dnat",
  28. "SDS": "true",
  29. "SERVICE_ACCOUNT": "istio-ingressgateway-service-account",
  30. "WORKLOAD_NAME": "istio-ingressgateway"
  31. },
  32. "userAgentBuildVersion": {
  33. "version": {
  34. "majorNumber": 1,
  35. "minorNumber": 15
  36. },
  37. "metadata": {
  38. "build.type": "RELEASE",
  39. "revision.sha": "f98b7e538920abc408fbc91c22a3b32bc854d9dc",
  40. "revision.status": "Clean",
  41. "ssl.version": "BoringSSL"
  42. }
  43. },
  44. },
  45. ...

Verifying connectivity to Istiod

Verifying connectivity to Istiod is a useful troubleshooting step. Every proxy container in the service mesh should be able to communicate with Istiod. This can be accomplished in a few simple steps:

  1. Create a sleep pod:

    1. $ kubectl create namespace foo
    2. $ kubectl apply -f <(istioctl kube-inject -f samples/sleep/sleep.yaml) -n foo
  2. Test connectivity to Istiod using curl. The following example invokes the v1 registration API using default Istiod configuration parameters and mutual TLS enabled:

    1. $ kubectl exec $(kubectl get pod -l app=sleep -n foo -o jsonpath={.items..metadata.name}) -c sleep -n foo -- curl -sS istiod.istio-system:15014/debug/endpointz

You should receive a response listing the “service” and “endpoint” for each service in the mesh.

What Envoy version is Istio using?

To find out the Envoy version used in deployment, you can exec into the container and query the server_info endpoint:

  1. $ kubectl exec -it productpage-v1-6b746f74dc-9stvs -c istio-proxy -n default -- pilot-agent request GET server_info --log_as_json | jq {version}
  2. {
  3. "version": "2d4ec97f3ac7b3256d060e1bb8aa6c415f5cef63/1.17.0/Clean/RELEASE/BoringSSL"
  4. }