This article includes updates for Smart ID 23.04.

The number of users is an indicator how many concurrent users we will have in the system. Card officers and registration officers typically work in the system multiple hours per day, while self-service users typically access it a couple of minutes at a time and quite rarely. There might be a peak of self-service users, for example when a lot of certificates expire at the same time or if an organizational change is done that require self-service interaction.

For authentication requests via the Digital Access Portal, in an enterprise scenario there is typically a peak in the morning and then a constant load during the day. This causes traffic on the Digital Access Portal and the Smart ID Messaging service.

A high number of concurrent users will cause load on the application servers. In that case, a load balancer setup with multiple nodes should be considered for the corresponding components.

Consider these questions about users and roles: 

• How many users do you expect to be managed in the system, and for which use case? For example, employees, contractors, visitors, operators and administrators.

• Which roles do these users have in the system, and how many users have each role? For example, end users that use Digital Access for authentication, Self-Service users, card officers, registration officers, administrators and helpdesk.

• Do you plan to manage connected devices in the system? In that case, how many, and for which use case?

Consider the following points:

• Which type of information will be stored: users, requests, cards, certificates, entitlements, devices etc. And how many records of each type will be stored at the same time?

• Logging and audit: Smart ID provides different kinds of audit information that is stored in the database, such as changes on any item that is managed, and protocols about authentication requests. Consider any background processes that might cause a large amount of changes per night, that end up in these protocols.

• Binary data, such as photos or pdfs, might consume a lot of space in the database. Consider the size of the binary objects to be stored.