• cOverview

• Summary

• Recommendations and Jiras

• Test results

• Resource Utilization Test 1,2,3,4

  • Service CPU Utilization

  • Memory Utilization

  • RDS CPU Utilization

  • RDS Database Connections

  • Average active sessions (AAS)

• Appendix

  • Methodology

This document contains the results of testing Sharing local instances(SLI) for MARC Source records. https://folio-org.atlassian.net/browse/PERF-1124 - Test Sharing Records' performance from member to central tenant

• The duration on the first three tenants (cs00000int_0001 to cs00000int_0003) for one virtual user process is approximately the same, with an average of about 6.1 to 6.7 seconds. The duration on the last tenant (cs00000int_0004) is around 0.9 seconds, which is 10 and 5 seconds faster than in Quesnelia, respectively. For 2 virtual users, the duration is about 6.7 seconds on cs00000int_0001 and 5.8 seconds on cs00000int_0002. For 3 and 4 virtual users, duration remains around 6.7 seconds for cs00000int_0001 to cs00000int_0003, except for cs00000int_0004 at 0.9 to 0.9 seconds.

• No memory leak is suspected for SLI modules.

• Maximum CPU utilization was about 16% on mod-pubsub-b and mod-inn-reach 1VU-14%, 2VU-15%, 3VU-14%, 4VU-13%.

• RDS CPU utilization was from 20% for 1VU and up to 40% for 4VU. That is three times less than in Quesnelia.

• All virtual user processes finished successfully (without errors). Error rate = 0.
  
  

Test 1 was run to get the number to evaluate the performance of the release versions

Test 1.1.[CSP1] was run to compare sunflower critical service patch performance. Ticket: https://folio-org.atlassian.net/browse/PERF-1185

• It's important to add parameters to the task definition to reduce local instance sharing time

  • "name": "JAVA_OPTS", "value": " -Dinventory.sharing.di.status.poll.interval.seconds=2"

Test 1. Virtual user working sequentially on each of the 4 tenants.

Test 2. Virtual users working in parallel on 2 tenants.

Test 3. Virtual users working parallel on 3 tenants

Test 4. Virtual users working parallel on 4 tenants