Database server
Where the failure starts.
What depends on what
Follow the
consequences.
Write down how a system is built: which servers and ports it uses, which outside services it calls, where it moves if the main site fails. Those links become a shared description, and they show what a change or an outage will hit before you learn it the hard way.
“What depends
on this server?”
Usually one engineer knows, and that engineer is on holiday. AverDesk links servers, running copies of services, the services themselves and the products people use into one map the whole team can read (a dependency graph).
When something fails, AverDesk walks that map and shows what stops next. The team keeps the map up to date on a diagram instead of in a table.
Link rulesHow rules work
Service instanceResource
Allowed linkYour organization decides
Limit per link typeSet for each type
No circlesThe circle is shown when a link is refused
One failure starts a chain.
Take the database server that went down at night. Here is how the trouble travels along the map.
Service instance
The copy of the service running on that server.
Employee product
The app your people use. The links lead straight to it.
Refuse the circle.
Explain why.
If A depends on B and B depends on A, nobody can tell where a failure starts. When a new link would close such a circle (a cycle), AverDesk refuses it and shows the exact path that closes it.
Try it below and see what the refusal looks like.
Link checkTry it
Service AService B
Existing path: Service B → Resource C → Service A
This link would close a circle.
Service A → Service B → Resource C → Service A
The map stays free of circles. Check the existing links before you add this one.
Your organization sets the rules.
Decide what can link to what and how many links of each kind are allowed. The map then follows the way your team actually thinks about its systems.
Find what’s forgotten.
The map lists things linked to nothing and things that never reach any product people use. A daily check reports anything newly forgotten, so the map does not quietly fall apart.
For the engineer. And for the service owner.
Engineers keep the links up to date. The night shift (in a NOC, a network operations center) follows the chain when something fails. The person responsible for a service finally sees what it runs on without asking anyone.