Check-in/check-out with parallel requests
- Olga Kondratenko
Test objectives
Perform check-in/check-out test with parallel requests to imitate FOLIO UX as much as it is possible. According to https://folio-org.atlassian.net/browse/PERF-220
General test conditions
Software versions
- mod-circulation-23.0.8
- mod-circulation-storage-14.0.0
- mod-inventory-18.1.6
- mod-inventory-storage- 23.0.2
- mod-authtoken-2.9.1
- mod-pubsub-2.5.0
- mod-patron-blocks-1.5.2
- mod-feesfines-17.1.1
- okapi- 4.13.2
Tests performed
Two types of tests have been modified and executed:
- Smoke test – to check functional readiness of tests.
- Baseline tests – to check load abilities.
Smoke test
Test duration: 1 loop;
Load type: ramp-up 1, 1 user.
Baseline tests
1st
Test duration: 30 minutes;
Load type: ramp-up -1 sec, user -1.
2nd
Test duration: 30 minutes;
Load type: ramp-up -80 sec, users -8.
Test results
Smoke test
The JMeter tests work properly, and the Load test can be performed
Baseline test 1st (1 user)
Performed baseline test with parallel requests has shown that the total number of requests for check-in increases by 19.15%, but at the same time the average response time increases by 28.18% in general, and for check-out total number of requests increases by 15.56%, but at the same time the average response time increases by 30.15%, in general, comparing to test with no parallel requests.
Without parallel requests
With parallel requests
Resources usage
There is almost no difference in resource usage between these tests performed for 1 user.
Baseline test 2nd (8 users)
Performed baseline test with parallel requests has shown that the total number of requests for check-in increases by 21.51%, but at the same time the average response time increases by 34.81% in general, and for check-out total number of requests increases by 13.43%, but at the same time the average response time increases by 34.10%, in general, comparing to test with no parallel requests.
Without parallel requests
With parallel requests
Resources usage
