Reg Proc and SyncData Performance Report
Overview
Identity life-cycle workflow underwent performance tuning in this release. Identity life-cycle consists of Registrations, Credential Processing and ID Authentication.
This report will cover only APIs for Registration Packet Upload & Sync Data from registration client.
This Performance Report provides a comprehensive analysis of the system’s responsiveness, reliability, and scalability for Registration client sync. It captures key performance metrics such as latency, throughput, resource utilization and memory consumption. This report is designed to help stakeholders understand how MOSIP v1.2.1.0 performs in real-world scenarios, its potential bottlenecks and recommendation for improvements.
For additional details on functional enhancements, refer to: Release Notes 1.2.1.0
Summary
A load test was executed to validate the performance and stability of the Registration Packet Upload and Sync Data APIs.
During the 9‑hour test window, the system successfully uploaded 700K+ registration packets, maintaining consistent throughput and keeping all API response times well below the 1‑second SLA target.
Resource Calculator is provided to help client countries estimate their hardware requirement to achieve similar performance at a larger scale.
No performance degradation or bottlenecks were observed throughout the test duration. Based on the results, this module is assessed to be stable, performant, and ready for release.
Test Environment
Deviation from the default
Camel and Reprocesser services were disabled to allow Registration Upload to be tested in isolation and not be hindered by packet and credential processing.
Software (Under Test)
Following ‘Images’ were under the scope of ‘Performance Testing’:
Modules Segregation
Image ID
Branch Name
Comments
mosipqa/registration-processor-stage-group-1:1.3.x
release-1.3.x
mosipqa/registration-processor-registration-status-service:1.3.x
release-1.3.x
mosipqa/registration-processor-registration-transaction-service:1.3.x
release-1.3.x
mosipqa/kernel-syncdata-service:1.3.x
release-1.3.x
mosipqa/kernel-keymanager-service:1.3.x
release-1.3.x
mosipqa/kernel-masterdata-service:1.3.x
release-1.3.x
mosipqa/kernel-auditmanager-service:1.3.x
release-1.3.x
mosipqa/kernel-auth-service:1.3.x
release-1.3.x
mosipid/kafka:3.2.1-debian-11-r9
Test Data
Performance data load has been populated before the run to ensure realistic results.
DB
Table Name
Number Of Records (Target)
Number Of Records (Tested with)
mosip_regproc
registration
1,000,000
610,765
mosip_master
registration_center
2,000
4,605
mosip_master
machine_master
5,000
5,035
Test Design
Test Duration: 9-Hours
Test Type: Load
Ramp Up: 2 mins
Total VUser Load: 30
Packets Uploaded: 730,000
Packet size: 0.4MB
Tools Used:
JMeter for user load simulation.
MOSIP Packet Creator for registration packet generation.
Workload Model
Scenario Name
Module Name
API Endpoint
SLA (ms)
Weightage/Load Distribution
Users
Throughput (TPS)
Target Volume
Sync And Upload New Registration Packet
Registration processor
registrationprocessor/v1/registrationstatus/syncV2
1000
90%
20
22.5
7,29,000
registrationprocessor/v1/packetreceiver/registrationpackets
1000
7,29,000
Sync Data From Server
Sync Data
v1/syncdata/authenticate/useridpwd
1000
5%
8
1.25
40,500
v1/syncdata/tpm/publickey/verify
1000
40,500
v1/syncdata/getCertificate?applicationId=${appId}&referenceId=${refId}
1000
40,500
v1/syncdata/userdetails?keyindex=${keyIndex}
1000
40,500
v1/syncdata/clientsettings?keyindex=${keyIndex}
1000
40,500
v1/syncdata/configs/${keyIndex}
1000
40,500
v1/syncdata/latestidschema?schemaVersion=${schemaVersion}
1000
40,500
v1/syncdata/getcacertificates
1000
40,500
Get Transaction Details From Reg Id
Registration processor
/registrationprocessor/v1/registrationtransaction/search/${regId}
1000
5%
2
1.25
40,500
Test Result
Performance test execution results
Application Name
Registration Processor & Sync Data
Test Duration
12/12/24 20:00PM - 13/12/24 5:00AM (9-Hours)
Number of users
30
Status
GREEN
Test Report
The 95th Line will be considered the primary Response Time metric while comparing with SLA.
Scenario Name
Transaction Name
API Endpoint
25 TPS & 30 VUsers
Date: 2025/12/12 (9 Hours)
# Samples
Min (ms)
Average (ms)
95% Line(ms)
Max (ms)
Error %
S01 Sync and Upload New Registration Packet
S01 T01 Sync Registration Packet Endpoint
registrationprocessor/v1/registrationstatus/syncV2
7,31,342
99.0
139.0
196.0
1155.0
0.1%
S01 T02 Upload Registration Packet Endpoint
registrationprocessor/v1/packetreceiver/registrationpackets
7,31,330
46.0
74.0
90.0
1139.0
0.0%
S02 Sync Data from Server
S02 T01 Auth Token Details Encrypted Based on Machine Key Endpoint
v1/syncdata/authenticate/useridpwd
39,358
60.0
76.9
93.0
201.0
0.0%
S02 T02 Public Key Verify Endpoint
v1/syncdata/tpm/publickey/verify
39,357
24.0
34.5
43.0
129.0
0.0%
S02 T03 Certificate Endpoint
v1/syncdata/getCertificate?applicationId=${appId}&referenceId=${refId}
39,357
24.0
37.8
48.0
107.0
0.0%
S02 T04 User Details Endpoint
v1/syncdata/userdetails?keyindex=${keyIndex}
39,356
50.0
66.3
80.0
200.0
0.0%
S02 T05 Client Settings Endpoint
v1/syncdata/clientsettings?keyindex=${keyIndex}
39,356
173.0
263.3
320.0
1006.0
0.0%
S02 T06 Get Configs Endpoint
v1/syncdata/configs/${keyIndex}
39,356
75.0
249.2
467.0
3057.0
0.0%
S02 T07 Get LatestId Schema Endpoint
v1/syncdata/latestidschema?schemaVersion=${schemaVersion}
39,355
31.0
46.6
58.0
233.0
0.0%
S02 T08 Get CaCertificates Endpoint
v1/syncdata/getcacertificates
39,354
518.0
613.0
688.0
1148.0
0.0%
S03 Get Transaction Details From Reg Id
S03 T01 Get Transaction Details From Reg Id Endpoint
/registrationprocessor/v1/registrationtransaction/search/${regId}
65,068
21.0
31.5
40.0
1042.0
0.0%
High Level Observations:
The response time was well under the 1 second target for all APIs.
We were able to perform 700K packets in 9 hours at the rate of 22 packets uploaded per second.
No degradation in response time was observed throughout the long test run.
Metrics
Response Time over Time
Observations:
Response time was consistent throughout the test with negligible variance.
Resource Utilization
For sake of brevity of the report, resource metrics are grouped by Namespace and all services in Namespace is displayed even if they did not actively participate in the test.
REGPROC
KERNEL
Observations:
Idgenerator service within Kernel namespace showed a brief spike (<8mins). This module is out of scope for this test. There was no measurable impact on overall system performance and throughput.
KEYMANAGER
Kafka
Observations:
One pod (out of 5) of Kafka service showed significantly noticeable dip in memory usage in the last 30 minutes of the test. But there was no measurable impact on overall system performance and throughput. The behaviour appears consistent with normal page‑cache activity, however reviewing memory allocation patterns may help refine resource sizing.
Resource Calculator
Find the 'Resource Calculator' here on this page.
Resource level configuration
Following configuration was used for the performance test.
NameSpace
Deployment
Resources
Java Option
No.of Pods
Limits
Requests
Min
Max
CPU(m)
Memory(Mi)
CPU(m)
Memory(Mi)
REGPROC
REGPROC-GROUP1
1000
1000
1000
1000
500
500
2
REGPROC
REGPROC-STATUS
700
1500
700
1500
950
950
2
REGPROC
REGPROC-TRANS
1000
5000
1000
5000
3750
3750
2
KERNEL
SYNCDATA
2000
2500
2000
2500
1750
1750
3
KEYMANAGER
KEYMANAGER
2000
6000
2000
6000
4250
4250
4
KERNEL
MASTERDATA
500
2250
500
2250
1500
1500
2
KERNEL
AUDITMANAGER
2000
5000
2000
5000
3250
3250
2
KERNEL
AUTHMANAGER
1000
5000
1000
5000
3150
3150
3
kafka
kafka
2000
5000
2000
5000
5
Performance Analysis
KPI
The performance test was conducted to ensure large amount of identity packets could be uploaded at a time. 700K+ packets were upload during 9-hour test run.
Sync Data activity was also simulated while concurrently while Registrations Upload was running. ‘Sync Data’ made up 5% of the traffic, Get Transaction Detail’ made another 5% and the remaining 90% was Registrations Upload.
Additionally, the hardware resources used to achieve this rate of transaction per second (TPS) has been shared via Resource Calculator document. It can be used to estimate resources required for higher processing as per specific needs of client countries.
Bottlenecks
Based on the defined scope and the results of this performance test, no major bottlenecks or performance‑limiting issues were observed. All services operated within acceptable resource boundaries, and all API response times remained well below the 1‑second SLA target.
Recommendation
The system demonstrated stable performance throughout the 9‑hour run, successfully processing 700K packets without any degradation. This confirms readiness for production‑scale workloads.
For capacity planning, it is recommended to use the Resource Calculator to estimate infrastructure requirements for higher throughput or larger population sizes.
Any deviations between the performance test environment, workload model and the default production configuration should be considered when planning resource allocation for real‑world deployments
.
Conclusion
MOSIP v1.2.1.0 is able to reliably upload 700,000+ registration packets in 9-hours. The Resource Calculator can be used to estimate the hardware requirements for higher upload capacity for larger population. No major performance degradation or bottlenecks were observed throughout the test duration, and this module is ready for release.
Last updated
Was this helpful?