Amazon Connect voice live stream

[zhangyuezhong]

I'm building a solution to download the Amazon Connect voice live stream, each call will have different streamARN and startFragment number.

AmazonKinesisVideo amazonKinesisVideo = (AmazonKinesisVideo) AmazonKinesisVideoClientBuilder.standard(). build();
String endpoint = amazonKinesisVideo.getDataEndpoint(new GetDataEndpointRequest().withAPIName(APIName.GET_MEDIA).withStreamName(streamName)).getDataEndpoint();

AmazonKinesisVideoMediaClientBuilder amazonKinesisVideoMediaClientBuilder = AmazonKinesisVideoMediaClientBuilder
		.standard()
		.withEndpointConfiguration(new AwsClientBuilder.EndpointConfiguration(endPoint, region.getName()))
		.withCredentials(awsCredentialsProvider);
AmazonKinesisVideoMedia amazonKinesisVideoMedia = amazonKinesisVideoMediaClientBuilder.build();

there will be 3400 calls over the day, I need to loop over and download them.

The question is how can i reuse amazonKinesisVideoMedia object (it is AmazonWebServiceClient), the endpoint could be different for each stream. (should i use Map to track the endpoint and client?)

it seems AmazonKinesisVideo and AmazonKinesisVideoMedia have separate HTTP Client and connection pool, they are not sharing the connection pool.

i can have single instance of AmazonKinesisVideo for my application and shutdown it before exit.

However, i have to create 3400 instance of AmazonKinesisVideoMedia ,

also are AmazonKinesisVideo and AmazonKinesisVideoMedia threadsafe ?