Duplicated nodes using http pagination

[stevesaner]

I am using Netbox as a source for oxidized. I am on the latest version of oxidized and here is the source config (oxidized is running on the same machine as netbox)

http:
url: http://127.0.0.1:8001/api/dcim/devices/?status=active&cf_oxidized=true
scheme: http
delimiter: !ruby/regexp /:/
headers:
Authorization: Token
map:
name: name
model: platform.name
ip: primary_ip4.address
group: role.name
insecure: true
secure: false
hosts_location: results
pagination: true
pagination_key_name: 'next'

This was working perfectly until I got to over 50 nodes. I then turned pagination on so that oxidized would make additional requests to netbox for the 2nd 50 nodes and so forth.

On the oxidized web interface, the first 50 nodes look just fine, but when I go to page 2, all of the nodes on page two are duplicated. I have a total of 56 nodes now defined in Netbox, but oxidized says there are 62 nodes. That's because 51 through 56 are there twice.

[ianbarrere]

This doesn't address the actual problem, but it may be a suitable workaround for you. You can provide a "limit" parameter in the Netbox URL to give you back more per page to avoid pagination completely. The URL I use in my source is this:

https://netbox.my.env/api/dcim/devices/?format=json&limit=1000&tag=oxidized

[stevesaner]

That does provide a workaround. Thanks for the suggestion.

[jseifeddine]

what about when we have >1000 devices

[rossnrachel]

I encounted the same issue. I have 2000 network devices, using http with pagination. There are about one thousand duplicated devices in Oxidized.

[github-actions]

This issue is stale because it has been open 90 days with no activity.