A tool for handling port-forwards to multiple services and across multiple clusters.
The setup is currently based on cargo until platform-specific binaries can be provided.
To install the latest version, run:
cargo install k8sfwdPlease note that the application internally relies on kubectl, so it needs to be present in your path.
If kubectl is not on your path, you may specify it via the --kubectl argument or
the KUBECTL_PATH environment variable.
Depending on your configuration, you'll be greeted with something along the lines of:
██╗░░██╗░█████╗░░██████╗░░░░░███████╗██╗░░░░░░░██╗██████╗
██║░██╔╝██╔══██╗██╔════╝░██╗░██╔════╝██║░░██╗░░██║██╔══██╗
█████═╝░╚█████╔╝╚█████╗░░╚═╝░█████╗░░╚██╗████╗██╔╝██║░░██║
██╔═██╗░██╔══██╗░╚═══██╗░██╗░██╔══╝░░░████╔═████║░██║░░██║
██║░╚██╗╚█████╔╝██████╔╝░╚═╝░██║░░░░░░╚██╔╝░╚██╔╝░██████╔╝
╚═╝░░╚═╝░╚════╝░╚═════╝░░░░░░╚═╝░░░░░░░╚═╝░░░╚═╝░░╚═════╝
k8s:fwd 0.1.0
Using kubectl version v1.24.12-dispatcher
Using config from .k8sfwd
Forwarding to the following targets:
#0 Items API (Staging)
target: service/foo.test-api
context: (default)
cluster: (default)
#1 Items API (Production)
target: pod/foo-59b58f5d68-6t6bh.test-api
context: (default)
cluster: production
Spawning child processes:
#0: Error from server (NotFound): pods "foo-59b58f5d68-6t6bh" not found
#0: Process exited with exit status: 1 - will retry in 5 sec
#1: Forwarding from 127.0.0.1:5012 -> 80
#1: Forwarding from 127.0.0.1:46737 -> 8080
#1: Forwarding from [::1]:5012 -> 80
#1: Forwarding from [::1]:46737 -> 8080
#0: Error from server (NotFound): pods "foo-59b58f5d68-6t6bh" not found
#0: Process exited with exit status: 1 - will retry in 5 sec
The configuration is provided as a YAML file. If no configuration file is specified when starting the application,
it will recursively look for a .k8sfwd file in the current directory hierarchy. If a file is specified on
program launch, that configuration is used instead.
See k8sfwd-example.yaml for an example.
---
version: 0.1.0
config:
# Optional: Number of seconds to wait before attempting to re-establish
# a broken connection.
retry_delay_sec: 5.0
targets:
- name: Test API (Staging) # Optional, for display purposes.
target: foo # The name of the resource to forward to.
type: service # Can be service, deployment or pod.
namespace: bar # The namespace of the resource.
context: null # Optional; will default to current context.
cluster: null # Optional; will default to current cluster.
ports:
- "5012:80" # Forward resource port 80 to local port 5012.
- "8080" # Forward resource port 8080 to random local port.
- name: Test API (Production)
target: foo-59b58f5d68-6t6bh
type: pod
namespace: bar
cluster: production
listen_addrs: # Select the listen addresses; defaults to `localhost`.
- "127.1.0.1"
ports:
- "5012:80"