AutoPilot for Apache Kafka provides operational and transactional monitoring for Apache Kafka, the open-source stream processing platform developed by the Apache Software Foundation, written in Scala and Java.
Apache Kafka is a unified, high-throughput, low-latency platform for handling real-time data feeds. It utilizes a massively scalable pub/sub message queue—designed as a distributed transaction log—as its storage layer. It is often used by Nastel clients as a transport mechanism for streaming data interconnected with other messaging and processing systems.
AutoPilot for Apache Kafka delivers a single-point-of-truth to track performance, latency, and logs, along with full message auditing and content surveillance capabilities. It provides complete message flow analytics, relating applications to the messages they publish to Kafka, and the applications that subscribe to those messages. Powerful network auto-discovery capabilities are included with matching visualization tools, as seen below. In the Viewlet below, solid lines (edges) represent send-to relationships; dotted lines represent an Acts-On relationship. Each edge has statistics showing average elapsed time and count.
Kafka network visualization showing an auto-discovered, pub-sub topology including senders, readers, and topics
AutoPilot for Apache Kafka offers powerful forensics to diagnose Kafka problems. Kafka performance and availability monitoring is accomplished via end-to-end stream monitoring and tracking of metrics from brokers, consumers, producers and Zookeeper, Kafka’s configuration service. AutoPilot for Apache Kafka examines the metrics collected for Kafka topics, producers, consumers and brokers while simultaneously offering deep-dive insight into the JVM itself.
- Make Kafka apps faster and more stable by:
- Optimizing Kafka apps
- Identifying latency and performance bottlenecks
- Identifying spots for data loss
- Improve quality and help debug apps by:
- Capturing Kafka exceptions
- Enabling easy message capture and profiling
- Generating message flow charts and topology displays
- Reduce MTTR (mean time to repair) by:
- Finding problems and bottlenecks
- Generating alerts based on user-defined conditions
- Providing anomaly detection in flows
- Identifying problems, data loss, and latencies
- Where is my message?
- Capture messages for problem identification
- Message flow analytics: see the longest and shortest paths
- Message Content Analytics
- Extract and summarize relevant business tokens. E.g., payment amounts (max, min, avg, etc.)
- Alerts based on business conditions
- SLAs and OLAs
- Single-pane-of-glass display for business flows over Kafka infrastructure
- Risk management and audit capabilities for relevant business transactions
- Auto discovery of end-to-end transactions spanning Kafka and other technologies such as IBM MQ
- Parsing of Kafka messages which are tokenized and utilized for analytics and transaction stitching
- SLA Monitoring, Analytics and Reporting
- Deep-dive monitoring of composite application components that include Kafka
- Proactive alerting and reduction in false alarms
…AutoPilot for Apache Kafka also utilizes Kafka internally as an integration technology for data transport.