Skip to main content

The Dependency map

Dependency map, under Service Configuration in the sidebar, draws services and how they rely on each other, and with the CMDB plugin the configuration items they run on, with the open work on each. Use it to see which services have incidents, problems or changes, during an incident to see who else to warn, and before a change to see what it can touch.

What it shows​

The map fills the page. It starts at the organization's business services and goes down to what they need. Open it from a service or a CI, or find one with the search field, and it is centred there and goes both ways: what it needs, and what needs it. A service's Map is on its Other relationships card (Relationships with the CMDB plugin); a CI's Open on the map is on its Around it card. Documentation at the top opens this page.

The controls float on the map: search and what to show on the left, how to draw it on the right.

ControlWhat it does
The search fieldA service by its code or name, or a CI by its name or an identifier. Picking one centres the map on it.
All servicesBack to the business services, when the map is centred on one.
Everything, With workWith work: the boxes with open work, and those it affects, stand out; the rest fades. The count is how many. Not shown for a past date.
DepthHow many steps from where it starts: 1, 2 (the default), 3 or All (up to 10).
What it needs, What needs it, BothWhich way it goes.
The arrowsLeft to right, or top down.
ImpactWith a node picked: everything trouble there would reach lights up, and the rest fades. On a service, that takes in its offerings; on an offering, not its service.
As ofA past date and time: the map as it was then (below). Clear it to come back to now.
CSVThe relationships on the map as a file, for a report or a spreadsheet.

Each box is a service or a CI, with its icon, code and name, type or class, and environment. Its left edge shows its health: green operational, blue maintenance, amber degraded, red outage, and dashed amber when it is at risk through something it relies on. CIs have squarer corners and a tinted background. Retired ones are faded. A + means more lies beyond: centre on it, or pick a larger depth. The legend at the bottom explains the lines, colours and flags.

Lines run from what relies to what it relies on. A solid line spreads trouble; a dashed one, such as Uses, doesn't. Point at a line to see its kind and note. Each of a service's offerings is joined to it by an Offering of line, which spreads trouble one way only: trouble on the service reaches its offerings, and trouble on an offering stays with the offering, never its service. That is the same as the service's health. Going down from a service, the map takes in its offerings too, and what they need.

Work on the map​

Flags on a box's top edge show the open work that names it, with how many of each:

FlagWork
Red shieldAn open major incident.
Red sirenAn open incident.
Amber bugAn open problem.
Blue hammerA change under way.
Blue calendarA change planned: waiting for approval, or scheduled.

A problem counts while it is open, known errors included. Point at the box to see each one's key, title and status.

An incident or major incident also marks everything that relies on what it names, through relationships that spread trouble, as Affected, with a red ring. So a database with an incident flags the applications that run on it and the business services above them. Only work you may see is shown.

As of a date​

Pick a moment under As of to see the map as it was then: for an audit, or to look back after an incident. It shows what was joined to what at that moment, rebuilt from the record of every relationship added and removed, and each service and CI with the name, type or class, status and environment its record had then. Services and CIs added since aren't on it, and relationships removed since are. A note at the top says which moment it shows; Back to now clears it.

Health and open work aren't kept per moment, so a map as of a date leaves them out: the boxes have no health edge and no flags. A CI's own health history is on its page (health).

Picking a node​

Click a box to see it on the side: its health and environment, the open work on it with a link to each, the incidents that affect it through what it relies on, how many others on the map trouble there could reach, and its relationships. Open goes to its page; Centre here redraws the map around it.

[!NOTE] The map shows what you may see. A service or CI in units you don't work in is a grey Hidden (no access) box without its name, and the map doesn't go past it. Each map is one organization's; nothing from another is ever on it.

A map shows at most 500 boxes. Past that it says so; centre on a smaller part, or pick a smaller depth.