Monitoring Apache NiFi with Prometheus and Grafana
Scraping NiFi metrics into Prometheus and building Grafana dashboards and alerts for queue depth, back pressure, processor throughput, JVM heap, repository usage and cluster node counts.
Apache NiFi operations and troubleshooting
Day-two work on a running NiFi: metrics, alerting, diagnostics, logging, stuck tasks, full disks, disconnected nodes and recovery.
Apache NiFi bulletins and alerting
Turning NiFi bulletins into alerts people act on: bulletin levels, the Site-to-Site bulletin reporting task, routing to Slack or email with PublishSlack and PutEmail, and avoiding alert floods from noisy processors.
Sending NiFi metrics and provenance out with reporting tasks
Using Site-to-Site reporting tasks to ship status, metrics, bulletins and provenance events to another NiFi or a downstream store, plus the Azure Log Analytics reporting task, and how much load they add.
Collecting Apache NiFi diagnostics and thread dumps
Gathering evidence before restarting a struggling NiFi: the diagnostics command and verbose mode, automatic diagnostics on shutdown, heap dumps and what to send a support team.
Finding stuck processors with the NiFi Long-Running Task Monitor
Detecting hung processor threads with the Long-Running Task Monitor: schedule and threshold settings, the thread dump cost, terminating tasks and the usual causes such as remote calls without timeouts.
Predicting back pressure with the NiFi analytics framework
Enabling queue predictions that warn before a connection hits its back pressure threshold: the regression model, snapshot frequency, the R-squared score threshold and using predictions in alerts.
Apache NiFi logging configuration and log shipping
Configuring logback for NiFi app, user and bootstrap logs, per-component log levels, deprecation logging, mapped diagnostic context and shipping logs to Loki or OpenSearch.
Recovering when the NiFi content repository fills the disk
What happens when the content repository runs out of space, how archive back pressure blocks writes, safe ways to free space, and the layout and threshold changes that stop it happening again.
Troubleshooting disconnected Apache NiFi cluster nodes
Diagnosing nodes that drop out of a NiFi cluster: missed heartbeats, flow mismatches that block rejoining, ZooKeeper or lease problems, certificate hostname errors after 2.12 and reconnect versus offload decisions.
Apache NiFi backup and recovery
Deciding what to back up for NiFi (flow.json.gz, configuration, keys, state and versioned flows in Git) versus what to rebuild, and restoring a node or cluster after loss without duplicating in-flight data.
Reading Apache NiFi JMX metrics through the REST API
Exposing selected JMX MBeans through the NiFi REST API: the allow-list filter pattern, required read permissions, bean name filters and feeding the results into existing monitoring.
