Versions Compared

Old Version 8

changes.mady.by.user xiangyu feng

Saved on

compared with

New Version 9

changes.mady.by.user xiangyu feng

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

In high concurrency scenarios, the CPU usage of the Flink cluster was reduced by 20% with no performance degradation in QPS by using proper retry strategy configuration.

initial-delay

200 ms

increment

100 ms

max-delay

700 ms


Baseline(128 Concurrency)

Experiment(128 Concurrency)

Retry Strategy

DEFAULT_RETRY_MILLIS=100 ms

collect-strategy.type

incremental-delay

Optimized retry strategy

Join QPS

47

48

JM CPU
Usage AVG

17.3

Image Modified

14.5

Image Modified

TM CPU
Usage AVG

9.0

7.2


Changed Public Interfaces

DeploymentOptions

Option Name

Default Value

Description

execution.interactive-client

false

Enabling this means the Flink client will be used in an interactive scenario. It should reuse all the necessary resources to improve interactive performance.

Notice: This option will only take effect when Flink is deployed in session-mode.

CollectOptions

Option Name

Default Value

Description

collect-strategy.type

fixed-delay

Config options for RetryStrategy in CollectResultFetcher.
The default RetryStrategy is FixedDelayCollectStrategy, which means CollectResultFetcher will keep fetch results from sinkoperator at fixed intervals.

collect-strategy.

fixed-delay.attempts

Integer.MAX_VALUE

Config options for FixedRetryStrategy, the strategy will retries at a fixed delay.

collect-strategy.fixed-delay.delay

100 ms

collect-strategy.

exponential-delay.initial-backoff

100 ms

Config options for ExponentialBackoffRetryStrategy

, the strategy will increase the delay exponentially with a cap

. The delay strategy starts at initial value and keeps increasing (multiplying by backoff multiplier) until maximum delay is reached.

collect-strategy.exponential-delay.max-backoff

1 s

collect-strategy.exponential-delay.

attempts

Integer.MAX_VALUE

collect-strategy.incremental-delay.initial-delay

100 ms

Config options for IncrementalDelayRetryStrategy, the strategy will increase the delay with a fixed increment.

backoff-multiplier

1.0

collect-strategy.

incremental-delay.increment

100 ms

collect-strategy.incremental-delay.max-delay

1 s

collect-strategy.incremental

exponential-delay.attempts

Integer.MAX_VALUE


Proposed Change

Overview

  • Reuse ClusterDescriptor when interacting with a dedicated session cluster
  • Reuse ClientHighAvailabilityServices for leader retrieval with the same session cluster
  • Reuse RestClusterClient when sending requests to the same session cluster
  • Reuse the HTTP connections between JDBC Driver to SQL Gateway and Flink Client to Flink Cluster with the same target address
  • Use ThreadLocal<ObjectMapper> instead of a global ObjectMapper instance
  • Use a configurable retry strategy in CollectResultFetcher to reduce unnecessary workload

...