-
Bug
-
Resolution: Done
-
Highest
-
El Alto Release
-
None
-
Frankfurt Sp1: 11/4 - 11/22
The blueprint processor had its liveness probe fail occasionally when the rate of incoming requests is higher than the rate it can execute.
This was causing disruption of service during that time where incoming requests rejected until the running requests finish running.
This situation was causing the endpoint “health-check” to fail to answer.
In order to replicate this behaviour with command executor, please take a look CBA I added and the following script which basically calls CDS multiple times.
#!/bin/bash N=15 echo "Will run N=${N} times the load test" echo "will sleep for 5 seconds before running,CTRL+C to cancel" #sleep 5for i in `seq 1 ${N}`; do (curl -X POST "http://<ip-to-reach-cds>:<port-to-reach-cds>/api/v1/execution-service/process" \ -H 'Accept: */*' \ -H 'Accept-Encoding: gzip, deflate' \ -H 'Authorization: Basic Y2NzZGthcHBzOmNjc2RrYXBwcw==' \ -H 'Content-Type: application/json' \ -d "{ \"commonHeader\": { \"originatorId\": \"System\", \"requestId\": \"123456\", \"subRequestId\": \"1234-12234\" }, \"actionIdentifiers\": { \"blueprintName\": \"remote_scripts\", \"blueprintVersion\": \"1.0.0\", \"actionName\": \"execute-remote-python\", \"mode\": \"sync\" }, \"payload\": { \"execute-remote-python-request\": { \"input\": \"RUN_NUM_${i}\" } } }";