Getting Started with Apache Flink

Explore Apache Flink's streaming-first architecture, its real streaming processing with batch as a special case, and key topics like history, architecture, and running a sample application.

Trace the history of Apache Flink from the Stratosphere project to incubator project, showing evolution from a Java API to a platform for batch, stream, graph processing, and machine learning.

Explore the architecture of Apache Flink, its layered components and data stream and data set APIs, the runtime with job graphs, and deployment options from local to yarn or cloud.

Discover Apache Flink's high-performance, low-latency features: exactly-once stateful computation, flexible time and session windows, fault-tolerant distributed snapshots, and memory-efficient stream and batch data processing.

Coordinate resources through the job client, job manager, and task manager. Submit the job via the job client; the job manager allocates resources and orchestrates tasks.

Explore how Flink's job managers coordinate task execution, scheduling, and fault-tolerant checkpoints with the Akka actor system enabling leader election and communication with task managers.

Explore how task managers serve as worker nodes that execute tasks in the JVM, allocate memory per task slot, balance parallelism across slots, and share TCP connections and heartbeat messages.

Accept the user program, transform it into a data flow, submit it to job manager, and return results after execution, as Flink partitions operators and streams for parallel, distributed processing.

download and install the Java Development Kit (JDK) from Oracle, accept the license, and install JDK 10.0.2 on Windows x64, optionally changing the path and creating a new folder.

Learn to set the path in environment variables by navigating from control panel to advanced system settings, add java home, and update the path accordingly.

Download Flink binaries to get started on Windows or Linux, and choose the binary compatible with your Hadoop version.

Install Flink on Windows by extracting files, navigating to the flink bin directory, and running start-cluster.bat with administrator rights. Confirm the local instance and web UI at localhost:8081.

Learn how to download VMware Workstation Player and Ubuntu ISO image, including navigating to official sites and initiating downloads.

Install VMware Workstation Player by double-clicking, accepting terms, choosing default settings, creating shortcuts in the Start Menu folder, and completing the installation.

Install Ubuntu inside VMware Workstation Player by creating a new virtual machine, configuring a username and password, and finishing setup; set memory to 2048 mb.

Start the window, enter the password, and sign in; if not started, go to player, then manage, then virtual machine settings, and set memory to 2048.

Install java on ubuntu: update package index, install the java runtime environment and jdk, add oracle java eight installer, configure default java with update-alternatives, set java_home in /etc/environment, and verify.

Learn to install Apache Flink on Ubuntu, prepare Java, download and extract Flink, set permissions, configure Flink home and environment, and start Flink in local mode with the dashboard.