Mojaloop Core test Harness

Run Mojaloop in your local machine using docker-compose without need for a Kubernetes cluster.

Pre-requisites

• git
• docker
• docker-compose

Starting mojaloop core services for a simple P2P transfer

Execute the following commands to run mojaloop in local machine

git clone https://github.com/mojaloop/ml-core-test-harness.git
cd ml-core-test-harness
docker-compose --profile all-services --profile ttk-provisioning --profile ttk-tests up

Wait for some time to get all the containers up and healthy. You can check the status of the containers using the command docker ps.

You should see the following output after some time. That means all your mojaloop services are up and test FSPs are onboarded successfully. Now you can run a P2P transfer.

┌───────────────────────────────────────────────────┐
│                      SUMMARY                      │
├───────────────────┬───────────────────────────────┤
│ Total assertions  │ 27                            │
├───────────────────┼───────────────────────────────┤
│ Passed assertions │ 27                            │
├───────────────────┼───────────────────────────────┤
│ Failed assertions │ 0                             │
├───────────────────┼───────────────────────────────┤
│ Total requests    │ 4                             │
├───────────────────┼───────────────────────────────┤
│ Total test cases  │ 1                             │
├───────────────────┼───────────────────────────────┤
│ Passed percentage │ 100.00%                       │
├───────────────────┼───────────────────────────────┤
│ Started time      │ Wed, 15 Jun 2022 17:02:28 GMT │
├───────────────────┼───────────────────────────────┤
│ Completed time    │ Wed, 15 Jun 2022 17:02:30 GMT │
├───────────────────┼───────────────────────────────┤
│ Runtime duration  │ 2398 ms                       │
└───────────────────┴───────────────────────────────┘

You can see all the test reports at http://localhost:9660/admin/reports and latest report should be available in reports/ folder.

Running P2P transfer again in a separate terminal session along with the running mojaloop

After all services been started, if you want to execute the P2P transfer from the command line again, use the following command in a separate terminal session.

docker-compose --project-name ttk-test-only --profile ttk-tests up --no-deps

_Note: This doesn't wait for any dependent services. You should make sure that all the services are up and healthy.

Running P2P transfer using testing toolkit web interface

• Open the URL http://localhost:9660 and go to Test Runner
• Goto on Collection Manager and click on the button Import Folder
• Select the folder docker/ml-testing-toolkit/test-cases/collections/tests and import
• Select p2p.json and click on the masked area on the right side of the screen
• Collection manager closes and click on Run button at the top right corner
• You should see all the tests passed
• You can explore the requests and responses by clicking on Edit button next to the test case

Running P2P transfer using testing toolkit mobile simulator

You can execute a transfer using the mobile simulator page where you can see two virtual mobile applications a sender and receiver.

By making a transfer using sender mobile application, you can see all the mojaloop requests and callbacks visually by means of a live sequence diagram.

http://localhost:9660/mobilesimulator

Profiles available

Profile Name	Description	Dependent Profiles
all-services	All mojaloop services including TTK	-
ttk-provisioning	For setting up mojaloop switch and onboard sample DFSPs	-
ttk-tests	TTK tests	-
debug	Debug utilities (kowl)	kafka
central-ledger	Central Ledger service	kafka
ml-api-adapter	ML API Adapter service	central-ledger
quoting-service	Quoting service	central-ledger
account-lookup-service	Account lookup service	central-ledger
discovery	Services used for discovery	-
agreement	Services used for agreement	-
transfer	Services used for transfer	-

Running various services with different profile combinations

Core services without provisioning

docker-compose --profile all-services up

Core services with debug utilities

docker-compose --profile all-services --profile debug up

Central ledger

docker-compose --profile central-ledger up

Quoting Service

docker-compose --profile quoting-service --profile central-ledger up

Note: We need to include central-ledger profile also here because its a dependency for quoting service

Account lookup service

docker-compose --profile account-lookup-service --profile central-ledger up

Note: We need to include central-ledger profile also here because its a dependency for account lookup service

ML API Adapter

docker-compose --profile ml-api-adapter --profile central-ledger up

Note: We need to include central-ledger profile also here because its a dependency for ml-api-adapter

Discovery

docker-compose --profile discovery up

Agreement

docker-compose --profile agreement up

Transfer

docker-compose --profile transfer up

Settlements

TODO: Add settlement related services

Bulk

TODO: Add bulk related services

Functional tests inside CICD

You can use this repo to run functional tests inside the CICD of a core service

The following commands can be added to the CICD pipeline

git clone --depth 1 --branch v0.0.2 https://github.com/mojaloop/ml-core-test-harness.git
cd ml-core-test-harness

docker-compose --project-name ttk-func --profile all-services --profile ttk-provisioning --profile ttk-tests up -d
bash wait-for-container.sh ttk-func-ttk-tests-1
docker logs ttk-func-ttk-tests-1 > ttk-tests-console.log
docker-compose -p ttk-func down
cat ttk-tests-console.log
ls reports/ttk-func-tests-report.html reports/ttk-provisioning-report.html

Performance Characterization

Running ALS with dependencies

docker compose --project-name ml-core -f docker-compose-perf.yml --profile als-test --profile ttk-provisioning-als up -d

Stop Services

docker compose --project-name ml-core -f docker-compose-perf.yml --profile als-test down -v

NOTE: -v argument is optional, and it will delete any volume data created by the monitoring docker compose

Running Services for Transfers characterization

docker compose --project-name ml-core -f docker-compose-perf.yml --profile transfers-test --profile 8dfsp --profile ttk-provisioning-transfers up -d

Stop Services

docker compose --project-name ml-core -f docker-compose-perf.yml --profile transfers-test --profile 8dfsp down -v

NOTE: -v argument is optional, and it will delete any volume data created by the monitoring docker compose

Running Services for Quotes characterization

docker compose --project-name ml-core -f docker-compose-perf.yml --profile quotes-test --profile 8dfsp --profile ttk-provisioning-quotes up -d

Stop Services

docker compose --project-name ml-core -f docker-compose-perf.yml --profile quotes-test --profile 8dfsp down -v

NOTE: -v argument is optional, and it will delete any volume data created by the monitoring docker compose

Running Services for Full E2E (Discovery+Agreement+Transfers) characterization

• Set ALS_SWITCH_ENDPOINT to "http://central-ledger:3001" in perf.env
• Set QS_SWITCH_ENDPOINT to "http://central-ledger:3001" in perf.env

docker compose --project-name ml-core -f docker-compose-perf.yml --profile all-services --profile 8dfsp --profile ttk-provisioning-e2e up -d

Stop Services

docker compose --project-name ml-core -f docker-compose-perf.yml --profile all-services --profile 8dfsp down -v

NOTE: -v argument is optional, and it will delete any volume data created by the monitoring docker compose

Running Services for SDK characterization

docker compose --project-name ml-core -f docker-compose-perf.yml --profile sdk-scheme-adapter up -d

Stop Services

docker compose --project-name ml-core -f docker-compose-perf.yml --profile sdk-scheme-adapter down -v

Setting up the Inbound/Outbound Server variables

• Go to perf.env and comment out the inboundSDK variables. You'll need to do the same and restart the docker-compose in order to change test suite.

Configuration for Transfers with batch support

• Set CENTRAL_LEDGER_POSITION_BATCH_REPLICAS to desired count in .env file
• Enable line CLEDG_KAFKA__EVENT_TYPE_ACTION_TOPIC_MAP__POSITION__PREPARE=topic-transfer-position-batch in perf.env file
• Set CENTRAL_LEDGER_VERSION to v17.2.0 or higher

Monitoring

Start Monitoring Services stack which uses:

• Prometheus for time series data store
• Grafana for visualization dashboards
• Node Exporter to instrument the Host machine
• CAdviser to instrument the Docker containers running on Host machine

docker compose --project-name monitoring -f docker-compose-monitoring.yml up -d

Stop Monitoring Services

docker compose --project-name monitoring --profile als-test --profile transfers-test -f docker-compose-monitoring.yml down -v

Start monitoring with account lookup service mysql exporter

docker compose --project-name monitoring --profile als-test -f docker-compose-monitoring.yml up -d

Start monitoring with central ledger mysql exporter

docker compose --project-name monitoring --profile transfers-test -f docker-compose-monitoring.yml up -d

or

docker compose --project-name monitoring --profile quotes-test -f docker-compose-monitoring.yml up -d

since the quoting service uses the central ledger database.

Start monitoring with all exporters

docker compose --project-name monitoring --profile als-test --profile quotes-test --profile transfers-test -f docker-compose-monitoring.yml up -d

NOTE: -v argument is optional, and it will delete any volume data created by the monitoring docker compose

TODO:

• add note about network being created by docker-compose-perf.yml, or it can be done manually.

Load Tests

K6 is being used to execute performance tests, with metrics being captured by Prometheus and displayed using Grafana.

Tests can be defined in the ./packages/k6-tests/scripts/test.js, refer to API load testing guide for more information.

Env configs are stored in the ./perf.env environment configuration file..

Note: Transfer testing and quote testing

Depending on the profile you started the performance docker compose with i.e --profile transfers-test --profile {2/4/8}dfsp You will need to edit K6_SCRIPT_FSPIOP_FSP_POOL json string in ./perf.env to contain 2/4/8 dfsps depending on your test. For reference here are the provisioned dfsps with an associated partyId available for use.

[
  {"partyId":19012345001,"fspId":"perffsp1","wsUrl":"ws://sim-perffsp1:3002"},
  {"partyId":19012345002,"fspId":"perffsp2","wsUrl":"ws://sim-perffsp2:3002"},
  {"partyId":19012345003,"fspId":"perffsp3","wsUrl":"ws://sim-perffsp3:3002"},
  {"partyId":19012345004,"fspId":"perffsp4","wsUrl":"ws://sim-perffsp4:3002"},
  {"partyId":19012345005,"fspId":"perffsp5","wsUrl":"ws://sim-perffsp5:3002"},
  {"partyId":19012345006,"fspId":"perffsp6","wsUrl":"ws://sim-perffsp6:3002"},
  {"partyId":19012345007,"fspId":"perffsp7","wsUrl":"ws://sim-perffsp7:3002"},
  {"partyId":19012345008,"fspId":"perffsp8","wsUrl":"ws://sim-perffsp8:3002"},
]

Start tests

env K6_SCRIPT_CONFIG_FILE_NAME=fspiopTransfers.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=fspiopTransfersUnidirectional.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=fspiopDiscovery.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=fspiopQuotes.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=fspiopE2E.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=inboundSDKDiscovery.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=inboundSDKQuotes.json docker compose --project-name load -f docker-compose-load.yml up
( or )
env K6_SCRIPT_CONFIG_FILE_NAME=inboundSDKTransfer.json docker compose --project-name load -f docker-compose-load.yml up

Cleanup tests

docker compose --project-name load -f docker-compose-load.yml down -v

SDK Security Overhead Testing

Regenerating Certificates

It's recommended that you do not trouble certificates and keys found in docker/security/. If you do need to for whatever reason these are the steps.

From the root ml-core-test-harness directory. Accept all defaults and enter y when prompted.

• cd docker/security/payer/jws/ && . keygen.sh && cd ../tls/ && . createSecrets.sh && cd ../../payee/jws && . keygen.sh && cd ../tls/ && . createSecrets.sh && cd ../../../../
• cp docker/security/payer/jws/publickey.cer docker/security/payee/jws/verification_keys/fspiopsimpayer.pem && cp docker/security/payee/jws/publickey.cer docker/security/payer/jws/verification_keys/fspiopsimpayee.pem
• cd docker/security/payer/tls/ && openssl ca -config openssl-clientca.cnf -policy signing_policy -extensions signing_req -out ../../payee/tls/dfsp_client_cert.pem -infiles ../../payee/tls/dfsp_client.csr && cp dfsp_server_cacert.pem ../../payee/tls/payer_server_cacert.pem && cd ../../../../
• cd docker/security/payee/tls/ && openssl ca -config openssl-clientca.cnf -policy signing_policy -extensions signing_req -out ../../payer/tls/dfsp_client_cert.pem -infiles ../../payer/tls/dfsp_client.csr && cp dfsp_server_cacert.pem ../../payer/tls/payee_server_cacert.pem && cd ../../../../

Here are more verbose hands on instructions of what above commands do.

• Run . keygen.sh and . createSecrets.sh in the /jws and /tls folders respectively for both payer and payee.
• Move payee/jws/publickey.cer to payer/jws/verification_keys/fspiopsimpayee.pem and move payer/jws/publickey.cer to payee/jws/verification_keys/fspiopsimpayer.pem
• Switch directories to docker/security/payer/tls/
• Run openssl ca -config openssl-clientca.cnf -policy signing_policy -extensions signing_req -out ../../payee/tls/dfsp_client_cert.pem -infiles ../../payee/tls/dfsp_client.csr
• Switch directories to docker/security/payee/tls/
• Run openssl ca -config openssl-clientca.cnf -policy signing_policy -extensions signing_req -out ../../payer/tls/dfsp_client_cert.pem -infiles ../../payer/tls/dfsp_client.csr
• Move each others dfsp_server_cacert.pem into each others folder and rename to payer_server_cacert.pem and payee_server_cacert.pem

Starting the Security Harness

• Run docker compose --project-name security -f docker-compose-security.yml --profile security-sdk-scheme-adapter up

Automate Load Tests

This section describes the process to automate capturing of grafana rendered dashboards after running the performance testing scenarios.

The main script that contains the logic for this is automate_perf.sh. Before running this script, the required variables are provided as environment variables that are defined in automate_perf.env. As this file contains login credentials, to avoid credential exposure a sample file called automate_perf_sample.env is available at the root level. Make a copy of this file, rename it to automate_perf.env and update the variable values.

Once automate_perf.env is ready, the next step is to make sure that the services for test harness and monitoring are up and running. The relevant docker-compose commands for these 2 steps are listed above in Performance Characterization section.

Once the required services are up and running, run automate_perf.sh from terminal. Once the script is completed successfully, a results folder is created at the main root level. In there another folder based on date is created and it creates subfolders for the different scenarios that are executed. The different dashboards that will be collected are specified in the script itself.

Run the script:

./automate_perf.sh

To capture results without running tests, use the following command

./automate_perf.sh -c -f <From Time in Milliseconds> -t <To time in Milliseconds>

Testing Performance of Remote Mojaloop Deployment

For executing performance test scenarios against a Mojaloop deployment, follow the steps below:

1. Set Environment Variables:

   • Set perf.override.env with the proper endpoints of the Mojaloop services.

2. Customize Configurations:

   • Edit the file docker/ml-testing-toolkit/test-cases/environments/remote-k8s-env.json to customize currencies and MSISDNs according to your requirements.

3. Run Simulators and TTK Provisioning:

   docker compose --project-name simulators -f docker-compose-perf.yml --profile 8dfsp --profile testing-toolkit --profile ttk-provisioning-remote-k8s --profile oracle up -d

4. Run Monitoring Services:

   docker compose --project-name monitoring --profile transfers-test -f docker-compose-monitoring.yml up -d

5. Execute Single Transfer Test Case:

   env K6_SCRIPT_CONFIG_FILE_NAME=fspiopSingleTransfer.json docker compose --project-name load -f docker-compose-load.yml up

6. Stop Services:

   docker compose --project-name simulators -f docker-compose-perf.yml --profile 8dfsp --profile testing-toolkit --profile ttk-provisioning-remote-k8s down -v
   docker compose --project-name monitoring --profile transfers-test -f docker-compose-monitoring.yml down -v

Note: The -v argument is optional and will delete any volume data created by the monitoring Docker Compose.

Helper scripts

The following helper scripts are available to allow easier execution of repetitive tasks.

• perf-test.sh - run various performance tests
• k8s-mojaloop-perf-tuning/patch.sh - patch k8s cluster to test with different releases and configurations

The easises way to use these scripts is to create bash aliases for them:

alias p='./k8s-mojaloop-perf-tuning/patch.sh'
alias t='./perf-test.sh'

Then use one of the following commands:

• t discovery - Test single account discovery
• t discoveries - Test account discoveries
• t discoveries rate - Test account discoveries with ramping rates
• t quote - Test single quote
• t quotes - Test quotes
• t quotes rate - Test quotes with ramping rates
• t transfer - Test single transfer
• t transfers - Test transfers
• t transfers rate - Test transfers with ramping rates
• t dqtrate - Test account discoveries, quotes and transfers in parallel with constant rates
• t e2e - Test multiple end to end
• t e2e single - Test single end to end
• t sim start - Start the simulators
• t sim stop - Stop the simulators
• t sim restart - Restart the simulators
• p audit - Configure direct events to Kafka without going through the sidecar, sending only audit and skipping others
• p direct - Configure direct events to Kafka without going through the sidecar
• p disabled - Configure no events to be produced
• p schema - Recreate central ledger schema object
• p init - Install RedPanda
• p - Restore baseline with events sent through the sidecar