Hostname Validation Bug

[coxley]

Appears #149 has a bug in hostname validation, at least something that might be unexpected.

This -is- configurable but I think it's unexpected as some people encode locations/environments in their FQDN. Maybe default to accepted dot? What do you think?

[jathanism]

I think the sane default is to only accept a hostname fragment.

I would prefer we:

• Highlight this fact the config and install documentation
• Use the error case "invalid name" to start a FAQ/user guide

[dowlingw]

+1 to allowing FQDN

As an example, my devices are named by site so the first label in the DNS name (eg: pe1) appears in every site. We operate a large MPLS L2VPN network so these devices would all be considered as part of the same "network" in nsot.

I don't think it would be too onerous to support both, or ditch the restriction entirely (looking forwards to nuts who don't use the DNS).

[dowlingw]

Whoops, ignore me - didn't fully understand the previous comment - can someone point me in the right direction for the configuration option? :)

[coxley]

Hey @dowlingw :)

If I understand correctly, you have devices like:

• pe1.site1.example.com
• pe1.site2.example.com
• etc...

If so then yeah, it's relevant to this ticket. It's possible but just not default. In nsot.conf.py, the parameter you want to tweak is DEVICE_NAME.

# Current default
DEVICE_NAME = re.compile(r"^([A-Za-z0-9][A-Za-z0-9\-]{0,61}[A-Za-z0-9]|[A-Za-z0-9])$")

# What I think you want
DEVICE_NAME = re.compile(r"^([A-Za-z0-9][A-Za-z0-9\-\.]{0,61}[A-Za-z0-9]|[A-Za-z0-9])$")

See: https://regex101.com/r/vO1xJ6/1

[jathanism]

Addressing this will fix #264