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
Copy file name to clipboardExpand all lines: development-docs/systemtests/io.strimzi.systemtest.performance.UserOperatorScalabilityPerformance.md
+2-2Lines changed: 2 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -22,8 +22,8 @@
22
22
23
23
| Step | Action | Result |
24
24
| - | - | - |
25
-
| 1. | Deploy Kafka cluster with User Operator configured with more resources to handle load and also non-default `STRIMZI_WORK_QUEUE_SIZE` set to 2048. | Kafka cluster with User Operator is deployed and ready. |
26
-
| 2. | For each configured load level (1000, 1500, 2000 existing users), create N KafkaUsers to establish the load. | N KafkaUsers are created and ready, establishing baseline load on the User Operator. |
25
+
| 1. | Deploy Kafka cluster with User Operator configured with more resources to handle load and also non-default `STRIMZI_WORK_QUEUE_SIZE` set to 4096. | Kafka cluster with User Operator is deployed and ready. |
26
+
| 2. | For each configured load level (1000, 2000, 3000 existing users), create N KafkaUsers to establish the load. | N KafkaUsers are created and ready, establishing baseline load on the User Operator. |
27
27
| 3. | Perform 100 individual user modifications sequentially, measuring the latency of each modification. | Each modification latency is recorded independently. |
28
28
| 4. | Calculate latency statistics: min, max, average, P50, P95, and P99 percentiles from the 100 measurements. | Statistical analysis shows how single-user modification latency degrades as system load (number of existing users) increases. |
29
29
| 5. | Clean up all users and persist latency metrics to user-operator report directory. | Namespace is cleaned, latency data is saved showing how responsiveness changes at different load levels. |
Copy file name to clipboardExpand all lines: systemtest/src/main/java/io/strimzi/systemtest/performance/gather/schedulers/BaseMetricsCollectionScheduler.java
+8Lines changed: 8 additions & 0 deletions
Original file line number
Diff line number
Diff line change
@@ -109,6 +109,14 @@ public void executeMetricsCollection() {
109
109
* @param unit the time unit of the initial delay and interval.
description = @Desc("This test measures user modification latency statistics under different load levels by performing multiple user modifications to understand how response time scales with system load."),
153
158
steps = {
154
-
@Step(value = "Deploy Kafka cluster with User Operator configured with more resources to handle load and also non-default `STRIMZI_WORK_QUEUE_SIZE` set to 2048.", expected = "Kafka cluster with User Operator is deployed and ready."),
155
-
@Step(value = "For each configured load level (1000, 1500, 2000 existing users), create N KafkaUsers to establish the load.", expected = "N KafkaUsers are created and ready, establishing baseline load on the User Operator."),
159
+
@Step(value = "Deploy Kafka cluster with User Operator configured with more resources to handle load and also non-default `STRIMZI_WORK_QUEUE_SIZE` set to 4096.", expected = "Kafka cluster with User Operator is deployed and ready."),
160
+
@Step(value = "For each configured load level (1000, 2000, 3000 existing users), create N KafkaUsers to establish the load.", expected = "N KafkaUsers are created and ready, establishing baseline load on the User Operator."),
156
161
@Step(value = "Perform 100 individual user modifications sequentially, measuring the latency of each modification.", expected = "Each modification latency is recorded independently."),
157
162
@Step(value = "Calculate latency statistics: min, max, average, P50, P95, and P99 percentiles from the 100 measurements.", expected = "Statistical analysis shows how single-user modification latency degrades as system load (number of existing users) increases."),
158
163
@Step(value = "Clean up all users and persist latency metrics to user-operator report directory.", expected = "Namespace is cleaned, latency data is saved showing how responsiveness changes at different load levels.")
0 commit comments