Failure Groups should provide indication when the list is too large to display

[rbev]

Describe the suggested improvement

Is your improvement related to a problem? Please describe.

If a ServiceControl instance has more than 200 distinct failure groups for a given grouping there has been a deliberate design decision to limit the list is truncated to the 200 most recently failed groups.

While this should be an unusual state for a system to be in, if it were to happen there are two main problems:

• No row count or total is shown in the UI so it is not possible to know that some failures are being hidden
• Inverting the sort specifier happens in the client so seeing older groups is simply not possible

Some sort of user feedback should be possible to know that they are not being shown a complete picture and have some action to take.

Describe the suggested solution

• Display an warning banner to indicate that there are failures hidden, and recommend archiving or retrying older failures to reduce the number of groups.

Describe alternatives you've considered

• Paging: while this would allow a user to see all the hidden records there is no evidence that this is a common enough use case to warrant adding paging to this interface.

Additional Context

Related PRs

• Move hardcoded rowlimit out of the storage implementation ServiceControl#5899