Planning Sites
How sites separate addresses, access and settings, and how many sites to create.
A site is a location your network runs in.
- Each site has its own map. Make a site cover the devices you want to see together: a utility whose fiber ring runs between buildings across a city can make the city one site to get a single map encompassing everything, and a company with offices in several cities can make each office a site to get separate maps for each office.
- Addresses are matched within a site. A subnet is identified by its CIDR, so two reports of
192.168.10.0/24on the same site are one subnet, and hosts on it merge by IP address. Hosts never merge across sites. - Access can be limited to a site. Users and API keys can be limited to specific sites, and credentials are assigned to sites or hosts.
- Each site has its own staleness window.
- A daemon reports to one site. You choose it when you create the daemon, and it can't move to another site, so create sites first.
The number of sites you can create depends on your plan. See Pricing. Create multiple sites for either reason below.
Do any of the networks use the same address ranges?
You don't need multiple sites for addresses. Each range becomes its own subnet.
Should some people or integrations see only part of what you document?
You don't need multiple sites for access.
Create the sites
Create the sites in Inventory > Sites. Then plan the daemons for each site in Planning Daemon Deployment.