Easy ReviewDOP-C02
Monitoring and Logging7

Managed Service for Apache Flink

Run managed Apache Flink applications for streaming analytics.

Este conteúdo de estudo ainda não foi traduzido, então aparece em inglês. Ajude a traduzir

Process streaming records continuously, recover application state, and send results to downstream destinations.

Stream processing workloads

Amazon Managed Service for Apache Flink runs applications that analyze streaming data, a continuous flow of records. Think of it as managed compute for processing records as they arrive.

How it works

  1. Write a Flink application that reads a source, such as a Kinesis streamA data stream is a continuous flow of data records. or an Amazon MSK topic. A source connectorA connector lets a Flink application read external data or write to an external location. reads external data.
  2. Apply transformations, such as filtering, enrichment, or aggregation over time. A sink connector writes the results to an external destination, such as Firehose.
  3. The service provisions compute and manages scaling and availability zone failover resilience. It can checkpointA checkpoint is a running application's backup used to recover from an unexpected disruption. a running application and restore it after failure.

Exam cues

  • If the question says aggregate streaming records over time, think Managed Service for Apache Flink, because an application can process records and send the results to Firehose.
  • If the question says explore a stream interactively, think Flink Studio, because it supports interactive queries.
  • If the question says recover from an unexpected failure, think checkpoints, because the application uses the newest checkpoint when checkpointing is enabled.

Remember

  • Default checkpointing is enabled: the interval is 60 seconds and the minimum pause is 5 seconds.
  • A checkpoint is a backup used for failure recovery. A snapshotA snapshot is a backup of application state. backs up application state for restoration.
  • One Kinesis Processing UnitA Kinesis Processing Unit is a unit of compute capacity for a managed Flink application. provides one vCPU, 4 GB of memory, and 50 GB of running application storage.

Trap: Firehose is not limited to delivery without transformation. It can transform data before delivery; use Flink when the application needs stream processing.

Process and deliver a stream
12Kinesis streamManaged FlinkFirehose
  1. A Flink application reads records from a Kinesis stream.

  2. The application processes records and can send results to Firehose.

Da documentação da AWS (11)
  • Managed Service for Apache Flink runs applications that process and analyze streaming data using Java, Scala, Python, or SQL.
  • A DataStream application reads a source such as Kinesis Data Streams or an Amazon MSK topic, applies transformations, and writes to a sink.
  • A Flink application can enrich records, aggregate data over time, and write results to a Firehose stream.
  • The service provisions compute and handles availability zone failover resilience, parallel computation, automatic scaling, checkpoints, and snapshots.
  • Flink Studio supports interactive stream queries, while a Flink application is suited to long-running streaming workloads.
  • With checkpointing enabled, a running application automatically creates checkpoints and uses the newest one to recover from failure.
  • A snapshot backs up application state and can be used to restore an application.
  • Default checkpointing is enabled with a 60000 millisecond interval and a minimum pause of 5000 milliseconds.
  • A Kinesis Processing Unit provides one vCPU, 4 GB of memory, and 50 GB of running application storage.
  • Amazon Data Firehose delivers streaming data to destinations and can transform data before delivery.
  • Amazon Managed Service for Apache Flink was previously named Amazon Kinesis Data Analytics for Apache Flink.

Flashcards

toque para virar
Resposta

·

Todos os 7 flashcards em lista

Which AWS service runs Apache Flink applications to process streaming data, rather than primarily delivering it?

Amazon Managed Service for Apache Flink. It runs stream processing applications; Firehose delivers data and can also transform it.

A team needs to aggregate Kinesis stream records over time and send results to Firehose. What can process the records?

Amazon Managed Service for Apache Flink. A Flink application can aggregate records and write results to Firehose.

A team wants to query a data stream interactively using SQL. Which Managed Service for Apache Flink option fits?

Flink Studio. It supports interactive queries of streams.

How does a Managed Service for Apache Flink application recover from an unexpected failure when checkpointing is enabled?

It uses the newest automatically created checkpoint.

A team wants to run a real-time log analytics application on Apache Flink without managing servers, scaling or checkpoints. Which service fits?

Amazon Managed Service for Apache Flink. It provisions the compute and handles scaling, checkpoints and snapshots for the application.

Does Amazon Data Firehose need to run an Apache Flink application to transform data before delivery?

No. Firehose can transform data before delivering it.

An exam answer mentions Amazon Kinesis Data Analytics for real-time analytics on a stream. What is that service called today?

Amazon Managed Service for Apache Flink.

Você concluiu todos os tópicos da DOP-C02!

Seu badge ficou dourado. Próximo passo: agendar a prova na AWS. Depois de passar, adicione-se ao Quem passou.

Compartilhar

Conte que você concluiu esta certificação

Concluí todos os tópicos da AWS Certified DevOps Engineer - Professional (DOP-C02) no Easy Review. Venha estudar também!