Introduction
Purpose
The purpose of this stress test plan is to evaluate the performance, stability, and scalability of the Call For Rent process under stress load conditions. The objective is to identify potential bottlenecks, weaknesses, and the maximum capacity of the application before failure or significant performance degradation occurs.
Objectives
- Confirm the flow itself is stable and covers the expected scenarion
- Validate the preliminary detected issue witn Accounting ledger
Environments / Infrastructure
Nikita's primary environmnet (shared): https://dev-nikita.life-connect.fr/
(Tech details of the Env)
Pipeline used for run: FLCS-3362_LifeConnect_handle_CFR
Run paramenets (full cleanup)
| Name | Value |
|---|---|
| CLEANUP_BEFORE | true |
| ENV_NAME | .env.nikita |
| RECORD | false |
Run paramenets (call 1st rent)
| Name | Value |
|---|---|
| CLEANUP_BEFORE | false |
| CMD_TO_RUN | call-rc-first-rent |
| ENV_NAME | .env.nikita |
| RECORD | true |
Run parameters (call regular rent)
| Name | Value |
|---|---|
| CLEANUP_BEFORE | false |
| CMD_TO_RUN | call-rc-regular-rent |
| ENV_NAME | .env.nikita |
| RECORD | true |
Data Set
Small size dataset: 100RC per 5 Orgs
Data set location: 1c3qMdnihRtKzKvMrwGPzaF2dcK2g1rSpaRc4gcESd1c
Scenari(s)
Provide full cycle of Test flow including:
- Full data cleanup/upload (w/o recording):
- cleanup all the collections (besides the initial company list)
- Call for 1-st rent (recodring)
- Call for regularall (recodring)
Collected run reports
The are 2 types of reports colected during the run:
- standard Artillery reports
- JFR (Java Flight Recorder) reports
Raw reports collected during the run are connected to test run itself and stored as bitbacket artifacts: raw report 1, raw report 2
As well as raw reports HTML-based version of Artillery could be view directly for:
Summary and conclusions
From the goals perspective run could be considered as successfull:
- Full run flow is configurable and stable
- The potensial issue with accounting was formally confirmed
As a next step it was recomended to pass the identical test for the dedicated (not shared) environment. The main assumption that need to be confirmed - the core reason of the issue is async nature of the accounting ledger data calculation.
And to have a relavant results the espessially Production-like environment need to be used.