refactor(inkless:s3): enable adaptive retry mode (#442)

This mode shows better performance under concurrent, high-troughput
scenarios by accommodating request rates to cope with the throughput
needed.

Author: jeqo

storage/inkless/src/main/java/io/aiven/inkless/storage_backend/s3/S3ClientBuilder.java

@@ -21,6 +21,7 @@
 
 import software.amazon.awssdk.auth.credentials.AwsCredentialsProvider;
 import software.amazon.awssdk.core.internal.http.loader.DefaultSdkHttpClientBuilder;
+import software.amazon.awssdk.core.retry.RetryMode;
 import software.amazon.awssdk.http.SdkHttpConfigurationOption;
 import software.amazon.awssdk.regions.Region;
 import software.amazon.awssdk.services.s3.S3Client;
@@ -61,6 +62,8 @@ static S3Client build(final Metrics metrics, final S3StorageConfig config) {
             c.addMetricPublisher(metricCollector);
             c.apiCallTimeout(config.apiCallTimeout());
             c.apiCallAttemptTimeout(config.apiCallAttemptTimeout());
+            // Optimizes for concurrency and high throughput use-cases by adjusting request rate dynamically
+            c.retryStrategy(RetryMode.ADAPTIVE_V2);
         });
         return s3ClientBuilder.build();
     }