A dashboard with an overview for Elasticsearch metrics:
- Cluster status (green/yellow/red)
- Nodes Statistics (master, data, ingest, etc)
- Operations time (indexing, refreshing, flushing)
- Index Statistics (documents, refreshes, flushes)
- HTTP connections and JVM Heap usage
- Search performance (queries and fetches, time, latency)
- Indexing performance (documents, status, latency)
- Cluster health and stats API (nodes, shards, docs, cache, storage, etc)
- Queues and rejections (index, search, write)
- Host metrics (file descriptors, bandwidth)
- Memory usage and garbage collector (heap, direct buffers, mapped buffers, etc)
- Indices segments (usage by type, total)
- Translog (operations, size)
- Feilddata cache (usage, evictions)
More dashboards for Netdata you can find here.
How to use
Follow these instructions to setup Elasticsearch monitoring in Netdata.
Please note that you need Netdata as an exporter for metrics. Plus, these labels are mandatory:
In your prometheus.yml it should look like this:
- job_name: netdata metrics_path: /api/v1/allmetrics?format=prometheus_all_hosts&source=raw relabel_configs: - source_labels: [__address__] regex: ^(.+)\.\w+:\d+ target_label: instance action: replace static_configs: - targets: [netdata.hostname.here:19999] labels: env: production group: applications source: newproject
WARNING: Without these labels, this dashboard won't be fully functioning.
OSSHelp Team, see https://oss.help