-
As per the use case, I created a trigger from AWS DynamoDB to call AWS Lambda.
-
The AWS Lambda will then call the lambda function to execute.
-
In this scenario I had noted down the timestamp of start of the execution of a program that pushes the data to the remote LAMBDA function via AWS dynamodb.
-
Latency: It is the time an application takes to be on wire. That means from the time the application is triggered from the clients' environemnt(whether UI or code) till the application ends.
-
I have taken 5 test cases with different input size.The input for each recod remains in same format.Each test pushds a row of data to the dyanamdb table. I captured the latency when the data was transferred from clients' code to update in AWS Dyanamodb database to triggering the AWS Lambda function.Below are the test case and corresponding results
| number of records inserted in DB | time taken to execute the whole process(in second) |
|---|---|
| 500 | 1.7 seconds |
| 1000 | 29.04 seconds |
| 2500 | 05 min 47.776 seconds |
| 5000 | 12 mins 38.903 seconds |
| 10000 | 26 mins 33.789 seconds |
-
As per the use case, I created a trigger from IBM cloudant to call IBM openwhisk.
-
The IBM openwhisk will then call the openwhisk function to execute.
-
In this scenario I noted down the timestamp of start of the execution of a program that pushes the data to the remote openwhisk function via IBM cloudant
-
I have taken 5 test cases with different input size.The input for each recod remains in same format.Each test pushds a row of data to the dyanamdb table. I captured the latency when the data was transferred from clients' code to update in IBM cloudant database to triggering the IBM openwhisk function.Below are the test case and corresponding results
| number of records inserted in DB | time taken to execute the whole process(in seconds) |
|---|---|
| 500 | 20.7 seconds |
| 1000 | 1 min 4.675 seconds |
| 2500 | 04 min 67.569 seconds |
| 5000 | 11 mins 35.656 seconds |
| 10000 | 30 mins 23.005 seconds |
Below is the graph comaprison.As per the chart we can see that IBM Openwhisk-cloudant combination is swift when it comes to small data but AWS lambda-dynamodb combination experience less latency when it comes to upload high amount of data:
