Closed ktbyers closed 5 years ago
Proposal on normalized values:
Bit | Value | Cisco string | Juniper string | Capability according to IEEE |
---|---|---|---|---|
0x001 | other |
O * |
Other |
Other |
0x002 | repeater |
P * |
Repeater |
Repeater |
0x004 | bridge |
B * |
Bridge * |
MAC Bridge component |
0x008 | wlan-access-point |
W * |
WLAN Access Point * |
802.11 Access Point (AP) |
0x010 | router |
R * |
Router * |
Router |
0x020 | telephone |
T * |
Telephone * |
Telephone |
0x040 | docsis-cable-device |
C * |
DOCSIS Cable Device |
DOCSIS cable device |
0x080 | station |
S * |
Station Only * |
Station Only |
0x100 | C-VLAN component | |||
0x200 | S-VLAN component | |||
0x400 | Two-port MAC Relay component |
* confirmed from sample output
sources:
/usr/sbin/lldpd
on a Junos device (75% sure the CLI output come from that binary)Yes, that looks good.
I assume we would use the "Value" column above as the normalized data.
We can look at our current unit test data (I glanced at it yesterday and IIRC it only had "B" and "R").
One item I noticed is Arista had "bridge" and "router" in all lower case (once again from memory so definitely double check).
Cisco "B" and "R" and abbreviated codes looked consistent across IOS, NX-OS, and IOS-XR platforms at least on our small sample size of data.
@jobec FYI, I am assuming you are writing the PR for this?
Mainly focusing on remote_system_capab and remote_system_enabled_capab
Adding @jobec I am open to more discussion on this.
I think the main requirements would be:
The normalization helper should probably go into base/helpers.py.
We would also have to agree on what values should be returned (i.e. what values we are going to standardize on).