The Finesse desktop calls the timestamponly version of the SystemInfo at a regular interval to calculate the client/server drift time. You will probably notice a log statement related to that right under the SystemInfo request.
Finesse relies on the SystemInfo Notification event to see if it needs to switch to the secondary Finesse server rather than polling the SystemInfo API itself.
As far as the speed of your API calls versus the the Finesse Desktop, are both the SPA and Finesse desktop on the same network? Is it on the same client machine? The Finesse Desktop doesn't have any special privileges where it gets higher priority. It uses the same REST APIs that are exposed for you to use in your on application.
Also, where is the latency?
Client -> Finesse server
Within the Finesse server while processing the request
Finesse -> Client
Our SPA is hosted in AWS. Once SPA is loaded in client browser it calls Finesse APIs to login, change state, etc (just like Finesse Desktop). So, the finesse API requests are being made from same network.
The latency issue is intermittent and happens from any agent desktop randomly. Latency seem to be between client and finesse server. From logs, we realized that, from the time request was made from SPA till the time Finesse logs shows first log entry of associated API request is about 15 seconds in some cases (we tracked times using requestid - sent in Finesse API request header).
We ran one test - where we ran Finesse Desktop and our SPA app side-by-side (using different agentid/deviceid, but same PG) and found out that Finesse Desktop didn't observe any latency, while our SPA did.
One suspicion we have is that Finesse Desktop and Finesse API are hosted on same machine (the finesse server) - so there is no CORS involved. Our SPA app is hosted in AWS (different domain than the finesse servers), so API request requires CORS to be cleared, before actual GET/PUT/POST request can be sent.
Any suggestions are greatly appreciated.