[BUG] Issue with long node names in hybrid provider scenarios

[jbemmel]

Lab:

---
groups:
  _auto_create: True
  nodes:
    members: [ verylongname1, verylongname2 ]
    device: frr
    provider: clab
  tor:
    members: [ r ]
    device: dellos10

links:
- r-verylongname1
- r-verylongname2

Result:

   ERROR 

  Root network namespace endpoint "verylongna_eth1" defined by multiple nodes [virbr1, virbr2].

Error executing sudo -E containerlab deploy --reconfigure -t clab-augment.yml:

The node name prefix needs to be unique in the first 10 characters. Possible solution: use a hex hash value of the node name, or the first 4 and last 4 characters

[ipspace]

I'm sorry for your headache, but I don't think this is worth fixing (but did you check how the 10-character limitation got there? 61d8a32 😜), particularly considering you could run Dell OS10 in vrnetlab container.

Also:

• Someone would have to check the code to verify we're never using the bridge interface names for any reason
• First N characters of a hash are not guaranteed to be unique.

We will need a generic solution, though, now that containerlab started using longer interface names for SR-SIM.