Skip to main content
Version: 2.20

Troubleshooting

Connecting to a cluster via VPN doesn't work.

There are a few different issues that could arise when working with a VPN. Please see the dedicated page on Telepresence and VPNs to learn more on how to fix these.

Connecting to a cluster hosted in a VM on the workstation doesn't work

The cluster probably has access to the host's network and gets confused when it is mapped by Telepresence. Please check the cluster in hosted vm for more details.

Volume mounts are not working on macOS

It's necessary to have sshfs installed in order for volume mounts to work correctly during intercepts. Lately there's been some issues using brew install sshfs a macOS workstation because the required component osxfuse (now named macfuse) isn't open source and hence, no longer supported. As a workaround, you can now use gromgit/fuse/sshfs-mac instead. Follow these steps:

  1. Remove old sshfs, macfuse, osxfuse using brew uninstall
  2. brew install --cask macfuse
  3. brew install gromgit/fuse/sshfs-mac
  4. brew link --overwrite sshfs-mac

Now sshfs -V shows you the correct version, e.g.:

$ sshfs -V
SSHFS version 2.10
FUSE library version: 2.9.9
fuse: no mount point
  1. Next, try a mount (or an intercept that performs a mount). It will fail because you need to give permission to “Benjamin Fleischer” to execute a kernel extension (a pop-up appears that takes you to the system preferences).
  2. Approve the needed permission
  3. Reboot your computer.

Volume mounts are not working on Linux

It's necessary to have sshfs installed in order for volume mounts to work correctly during intercepts.

After you've installed sshfs, if mounts still aren't working:

  1. Uncomment user_allow_other in /etc/fuse.conf
  2. Add your user to the "fuse" group with: sudo usermod -a -G fuse <your username>
  3. Restart your computer after uncommenting user_allow_other

Distributed tracing

Telepresence is a complex piece of software with components running locally on your laptop and remotely in a distributed kubernetes environment. As such, troubleshooting investigations require tools that can give users, cluster admins, and maintainers a broad view of what these distributed components are doing. In order to facilitate such investigations, telepresence >= 2.7.0 includes distributed tracing functionality via OpenTelemetry Tracing is controlled via a grpcPort flag under the tracing configuration of your values.yaml. It is enabled by default and can be disabled by setting grpcPort to 0, or tracing to an empty object:

tracing: {}

If tracing is configured, the traffic manager and traffic agents will open a GRPC server under the port given, from which telepresence clients will be able to gather trace data. To collect trace data, ensure you're connected to the cluster, perform whatever operation you'd like to debug and then run gather-traces immediately after:

Terminal
$ telepresence gather-traces

This command will gather traces from both the cloud and local components of telepresence and output them into a file called traces.gz in your current working directory:

Terminal
$ file traces.gz
traces.gz: gzip compressed data, original size modulo 2^32 158255

Please do not try to open or uncompress this file, as it contains binary trace data. Instead, you can use the upload-traces command built into telepresence to send it to an OpenTelemetry collector for ingestion:

Terminal
$ telepresence upload-traces traces.gz $OTLP_GRPC_ENDPOINT

Once that's been done, the traces will be visible via whatever means your usual collector allows. For example, this is what they look like when loaded into Jaeger's OTLP API:

Jaeger Interface

Note: The host and port provided for the OTLP_GRPC_ENDPOINT must accept OTLP formatted spans (instead of e.g. Jaeger or Zipkin specific spans) via a GRPC API (instead of the HTTP API that is also available in some collectors) Note: Since traces are not automatically shipped to the backend by telepresence, they are stored in memory. Hence, to avoid running telepresence components out of memory, only the last 10MB of trace data are available for export.

No Sidecar Injected in GKE private clusters

An attempt to telepresence intercept results in a timeout, and upon examination of the pods (kubectl get pods) it's discovered that the intercept command did not inject a sidecar into the workload's pods:

$ kubectl get pod
NAME READY STATUS RESTARTS AGE
echo-easy-7f6d54cff8-rz44k 1/1 Running 0 5m5s

$ telepresence intercept echo-easy -p 8080
telepresence: error: connector.CreateIntercept: request timed out while waiting for agent echo-easy.default to arrive
$ kubectl get pod
NAME READY STATUS RESTARTS AGE
echo-easy-d8dc4cc7c-27567 1/1 Running 0 2m9s

# Notice how 1/1 containers are ready.

If this is occurring in a GKE cluster with private networking enabled, it is likely due to firewall rules blocking the Traffic Manager's webhook injector from the API server. To fix this, add a firewall rule allowing your cluster's master nodes to access TCP port 443 in your cluster's pods, or change the port number that Telepresence is using for the agent injector by providing the number of an allowed port using the Helm chart value agentInjector.webhook.port. Please refer to the telepresence install instructions or the GCP docs for information to resolve this.

Injected init-container doesn't function properly

The init-container is injected to insert iptables rules that redirects port numbers from the app container to the traffic-agent sidecar. This is necessary when the service's targetPort is numeric. It requires elevated privileges (NET_ADMIN capabilities), and the inserted rules may get overridden by iptables rules inserted by other vendors, such as Istio or Linkerd.

Injection of the init-container can often be avoided by using a targetPort name instead of a number, and ensure that the corresponding container's containerPort is also named. This example uses the name "http", but any valid name will do:

apiVersion: v1
kind: Pod
metadata:
...
spec:
containers:
- ports:
- name: http
containerPort: 8080
---
apiVersion: v1
kind: Service
metadata:
...
spec:
ports:
- port: 80
targetPort: http

Telepresence's mutating webhook will refrain from injecting an init-container when the targetPort is a name. Instead, it will do the following during the injection of the traffic-agent:

  1. Rename the designated container's port by prefixing it (i.e., containerPort: http becomes containerPort: tm-http).
  2. Let the container port of our injected traffic-agent use the original name (i.e., containerPort: http).

Kubernetes takes care of the rest and will now associate the service's targetPort with our traffic-agent's containerPort.

Important note

If the service is "headless" (using ClusterIP: None), then using named ports won't help because the targetPort will not get remapped. A headless service will always require the init-container.

Error connecting to GKE or EKS cluster

GKE and EKS require a plugin that utilizes their resepective IAM providers. You will need to install the gke or eks plugins for Telepresence to connect to your cluster.

too many files open error when running telepresence connect on Linux

If telepresence connect on linux fails with a message in the logs too many files open, then check if fs.inotify.max_user_instances is set too low. Check the current settings with sysctl fs.notify.max_user_instances and increase it temporarily with sudo sysctl -w fs.inotify.max_user_instances=512. For more information about permanently increasing it see Kernel inotify watch limit reached.

Connected to cluster via VPN but IPs don't resolve

If telepresence connect succeeds, but you find yourself unable to reach services on your cluster, a routing conflict may be to blame. This frequently happens when connecting to a VPN at the same time as telepresence, as often VPN clients may add routes that conflict with those added by telepresence. To debug this, pick an IP address in the cluster and get its route information. In this case, we'll get the route for 100.124.150.45, and discover that it's running through a tailscale device.

Terminal
$ ip route get 100.124.150.45
100.64.2.3 dev tailscale0 table 52 src 100.111.250.89 uid 0

This will tell you which device the traffic is being routed through. As a rule, if the traffic is not being routed by the telepresence device, your VPN may need to be reconfigured, as its routing configuration is conflicting with telepresence. One way to determine if this is the case is to run telepresence quit -s, check the route for an IP in the cluster (see commands above), run telepresence connect, and re-run the commands to see if the output changes. If it doesn't change, that means telepresence is unable to override your VPN routes, and your VPN may need to be reconfigured. Talk to your network admins to configure it such that clients do not add routes that conflict with the pod and service CIDRs of the clusters. How this will be done will vary depending on the VPN provider. Future versions of telepresence will be smarter about informing you of such conflicts upon connection.