How-to recipes

Task-shaped guides for operating LinkMesh. Each one is self-contained — start wherever your problem is.

New to LinkMesh? Start with the Quickstart and Onboard your first log instead, then come back here.

Build and shape pipelines

Build your first pipeline — the chain editor end-to-end: add processors, dry-run, attach to a route
Preview a processor chain — the dry-run workspace in depth
Create a custom processor — define your own processor template when the built-in library doesn’t fit
Create a custom source or destination — write a receiver’s or exporter’s settings as YAML when the guided form doesn’t offer what you need
Add a collector extension — declare an authenticator, a storage or any other extension your collector contains, and reference it with ${ext:<label>}
Capture live samples — tap a live route and preview against real records
Drop noisy logs — filter out health-check probes and other low-value noise
Filter records on a route — route input filters, with an examples cookbook
Mask PII — redact sensitive values on the collector, before anything leaves the host
Generate test telemetry — produce traffic to verify a pipeline without waiting for real data

Enrol a collector — the general enrollment path
Onboard Grafana Alloy (remotecfg)
Onboard otelcol-contrib (OpAMP)
Install the agent (optional) — the edge connector, and when you actually need it
Manage collector groups — configure many collectors as one
Upgrade collectors
Resume a log file after a restart — keep your place in a file across a restart, upgrade or pod replacement

Install from APT or YUM — signed repository, architectures, and the signing-key fingerprint
Run behind a reverse proxy
Deploy with MongoDB
Run a highly-available deployment
Upgrade LinkMesh
Monitor the server
Set up alerting — notify on collector and pipeline health
Generate a support bundle — a redacted snapshot to attach to a support request

Something not working? Troubleshooting is organised by symptom.