Napalm LLDP neighbors works through the REST API, but fails intermittently in the UI

[danielc839]

Hello,

I am trying to figure out an issue with Napalm in our Netbox 3.0.2 deployment (netbox-docker 1.3.1). When I try to get LLDP neighbors via the API (e.g. https://acme.com/api/dcim/devices/2345/napalm/?method=get_lldp_neighbors), the data is returned successfully. However, if I try to get this data through the Netbox UI (e.g. https://acme.com/dcim/devices/2345/lldp-neighbors/), then the LLDP columns ("LLDP device" and "LLDP interface") are blank.

Looking at the logs, nothing seems wrong. These are the logs that show up when I reload the Netbox UI for LLDP neighbors:

netbox_1         | 204.186.63.30 - - [07/Dec/2021:20:15:34 +0000] "GET /dcim/devices/2345/lldp-neighbors/ HTTP/1.1" 200 131750 "-" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.93 Safari/537.36"
netbox_1         | 204.186.63.30 - - [07/Dec/2021:20:15:47 +0000] "GET /api/dcim/devices/2345/napalm/?method=get_lldp_neighbors_detail HTTP/1.1" 200 88 "https://netbox-dev.engr.ptd.net/dcim/devices/2345/lldp-neighbors/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.93 Safari/537.36"

And here is the request to the same device via the API:

netbox_1         | 204.186.63.30 - - [07/Dec/2021:20:17:45 +0000] "GET /api/dcim/devices/2345/napalm/?method=get_lldp_neighbors HTTP/1.1" 200 12511 "-" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.93 Safari/537.36"

I tried this with several NCS5508s (Cisco IOS XR), the issue seems to occur randomly. The same device works and then later does not work. EDIT: it actually seems like the issue happens fairly consistently on the same devices, though I believe I have seen devices not work, and then work. The response to the requests takes roughly 10 seconds.

Thinking this might be a timeout issue, I tried bumping up the Napalm timeout variable, and also setting a timeout of 60 seconds in nginx-unit.json. Neither worked.

Are there any other logs I could look at, or any other setting I could change for this?

Thank you in advance.