You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jan 28, 2022. It is now read-only.
Having added simulated latency to the mock api used for load-testing we see that as load increases the operator reaches a point where it cannot keep up and the queue keeps increasing
Below graph with the load test with latency ranges for GET (50-200ms) and other requests (200-1000ms) that ramps up from 0 to 25 users and remains constant at 25 users.
We see that the work load in the queue has increased and on an average waits in the queue for 10s and the queue depth keeps increases.
Graph of the time spent in the queue for a run
Graph of Controller queue depth
The text was updated successfully, but these errors were encountered:
creddy123
changed the title
Multiple reconcilers will improve operator performance
Operator cannot keep up with incoming load
Jan 29, 2020
Having added simulated latency to the mock api used for load-testing we see that as load increases the operator reaches a point where it cannot keep up and the queue keeps increasing
Below graph with the load test with latency ranges for GET (50-200ms) and other requests (200-1000ms) that ramps up from 0 to 25 users and remains constant at 25 users.
We see that the work load in the queue has increased and on an average waits in the queue for 10s and the queue depth keeps increases.
Graph of the time spent in the queue for a run
Graph of Controller queue depth
The text was updated successfully, but these errors were encountered: