As we do now support VRFs a new top-level op-mode command similar to ping and traceroute is required to support to source an SSH connection from a given VRF.
The command style should be inspired by Cisco IOS.
Description
Description
Details
Details
- Difficulty level
- Normal (likely a few hours)
- Version
- 1.3-rolling
- Why the issue appeared?
- Will be filled on close
- Is it a breaking change?
- Unspecified (possibly destroys the router)
- Issue type
- Feature (new functionality)
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Open | BUG | c-po | T2762 VRF: when SSHd is VRF bound all commands are executed in VRF context | ||
| Open | BUG | None | T2781 Create op-mode top-level wrapper for ssh command | ||
| Open | FEATURE REQUEST | None | T3915 Create op-mode top-level wrapper for ssh/scp command -VyOS 1.4 |
Event Timeline
Comment Actions
@c-po What do you propose?
ssh [email protected] vrf foo
Interpret to..
ip vrf exec foo ssh [email protected]
Comment Actions
Besides the command proposal there is a common issue with how we build op-mode commands. The best example is the ping op-mode command where and artifical node.tag folder is created which links (ln -s) back to itself.
Currently there is no way besides an external helper script which acts as completion helper to create the tab-completion commands for ping, traceroute or now ssh command.
Thus there are two options
a) it is at it is - write a ssh-completion helper which just shows the options without any additional help string (see ping as reference)
b) find a new and better way to implement op-mode commands which can have recursive options (linked back to the root nodes so we can either call: ssh user@host ipv4 vrf bar or ssh user@host vrf bar ipv4 without implementing the <node> structure twice for vrf on top and the ipv4 on top.