Exchange Topology
What this screen is for
The topology is a map of the business relationships your tenant does cryptographic work in. Your applications stand on the left, the counterparties they exchange with on the right, and each line between them is one exchange context, coloured by the sensitivity of the data it carries. Selecting a line shows who operates in that relationship, with which capabilities and which keys, and what happened there recently; selecting an application or a counterparty shows everything it is involved in.
The screen calls the lines exchanges, and the counts in its header say exchanges too. What it draws is the exchange context, one line per context, whether or not a cryptographic exchange has been composed on it. The map changes nothing by itself: it reads, and it offers two actions on a context and one link out.
What must exist first
This screen is for tenants that compose authorization by hand. It is not offered to SIMPLE tenants.
Before you work here:
- Exchange contexts. The map is empty until the tenant has at least one, and the empty state points you to the Cryptographic Exchanges screen to create the first.
- The permission to list exchange contexts, which shows the sidebar entry and loads the lines. Without it the screen reports the refusal.
- The permission to list applications. The map draws a context only under an application it could read, so without this permission every line is dropped and the screen says Nothing to map yet although contexts exist.
- For the detail panel, the permissions to list capability grants, actors and keys, and to read the tenant's crypto and exchange-context audit trails. Lacking the grant or actor permission removes those sections from the panel without a notice; lacking the key inventory shows Inventory unavailable, and lacking an audit permission is stated in the activity section in those words.
The map reads at most 200 contexts, the most recently created first, and at most 200 applications, the first by name, and draws nothing for what lies beyond either limit.
Decisions you make here
Reading the map
Every context of the tenant is drawn, whatever its status. A suspended one is dimmed and its flow animation stops; a closed one and one pending activation are drawn exactly like an active one, and only its detail says otherwise. A context never leaves the map when the exchange composed on it is revoked or expires.
Counterparties are grouped by the name the contexts record, ignoring case and spacing, so VISA and visa are one node. A context that names no counterparty joins the node No counterparty registered, unless its direction is INTERNAL, in which case it is drawn as a loop on its own application.
The colour is the context's sensitivity: CONFIDENTIAL is Critical, RESTRICTED is High, INTERNAL is Medium and PUBLIC is Low. The arrow on the line is its direction.
Filtering
The sensitivity and direction chips fade lines out and show how many lines each value has. Nothing is removed and the header counts do not change; the filter is a way to look, not a query. It resets when you leave the screen.
Selecting a line
The panel shows the context's business purpose, sensitivity, direction and status, and then what the active capability grants scoped to that context imply: the actors they name, the capabilities they enable, the constraint policy each grant enforces and the keys they cover, each key with its algorithm and status from the inventory. The constraint policy is listed per grant, with the version the grant pinned and its limits, because two grants in one context can be constrained differently; a grant without one says No constraint policy. A context whose exchange is still a draft, or whose grants have been revoked, shows none of these sections.
Below them, the activity in that context: its operations and its administrative changes, newest first, the most recent page refreshed every thirty seconds while the tab is focused. All shows the usage rows and the administrative changes to the context together; Administration and Usage show one plane. Audit details widens each row to the actor, by the name the record carries and its identifier, the counterparty and direction, the grant, the constraint evaluations the operation passed or failed and, for a re-encryption or re-signature, the source and target keys. A requested key is shown only when it differs from the key the operation used.
Each of the two states how many of its entries are loaded out of how many it holds, Load more reads the next ones until the panel says that history is complete, and on All the panel states the date the merged list is complete back to. View in Audit Log opens the Audit Log narrowed to the context over the last 365 days.
Selecting an application or a counterparty
The panel shows how many contexts it is involved in, a breakdown of their sensitivities, the application's risk level where it has one and otherwise the worst sensitivity among its contexts, and the list of its contexts to click through to. The No counterparty registered node works the same way.
Suspending and resuming a context
On an active context the panel offers Suspend, and on a suspended one Resume. These are the same actions the Exchange Contexts screen offers, with the same effects and the same permissions: suspending refuses every operation bound to the context at once and needs a written reason, which is recorded in the audit trail against your account; resuming needs none. The buttons are offered whether or not you hold the permission, and the server refuses when you do not. The context's own page describes what a suspension does and does not stop:
Suspending the context does not change the status of a cryptographic exchange composed on it; that exchange keeps reading ACTIVE on its list.
Edit
Edit opens the context on the Exchange Contexts screen, on its Edit tab, where its display name, description, counterparty, business purpose and sensitivity are changed. Without the permission to update exchange contexts it opens on the Info tab. If the context cannot be read, the screen says why. The cryptographic exchanges composed on the context are edited on their own screen.
How it relates to other screens
Each line is an exchange context, and everything a line shows is decided on that context's own screen:
The actors, capabilities and keys in a line's detail are read from the capability grants scoped to that context. A grant reaches the map only once it is active, which for an exchange means once it is activated:
The applications on the left are your applications, and the risk level in their detail is the one chosen when the application was created:
The activity in a line's detail is a window onto the same trail the Audit Log screen searches in full: