Grafana Tempo — a high-scale, cost-efficient distributed tracing backend with OTLP ingestion and local block storage. Query traces from Grafana. Deployed as a single host-networked Nomad service.
Needs nomad-pack on PATH. The script only adds the nomploy registry and runs this pack.
Source ↗ Project ↗ ★ 5.5k ⚑ Report an issue
Save as values.hcl, edit, then run:
# The name of the Nomad job.
job_name = "tempo"
# The Nomad namespace to deploy into.
namespace = "default"
# The datacenters to deploy to.
datacenters = ["*"]
# The Grafana Tempo container image. Pin a tag in production.
image = "grafana/tempo:latest"
# HTTP port for the Tempo API / query frontend (server.http_listen_port).
port = 3200
# OTLP gRPC ingest port.
otlp_grpc_port = 4317
# OTLP HTTP ingest port.
otlp_http_port = 4318
# How long to keep trace blocks before compaction removes them.
block_retention = "336h"
# Named volume mounted at /var/tempo (WAL + trace blocks).
data_volume = "tempo_data"
# Placement constraints. On a nomploy cluster: attribute = "$${meta.nomploy_control_plane}", operator = "=", value = "true".
constraints = []
# The task resources.
resources = {
cpu = 300
memory = 256
}
| Name | Type | Default | Description |
|---|---|---|---|
| job_name | string | "tempo" | The name of the Nomad job. |
| namespace | string | "default" | The Nomad namespace to deploy into. |
| datacenters | list | ["*"] | The datacenters to deploy to. |
| image | string | "grafana/tempo:latest" | The Grafana Tempo container image. Pin a tag in production. |
| port | number | 3200 | HTTP port for the Tempo API / query frontend (server.http_listen_port). |
| otlp_grpc_port | number | 4317 | OTLP gRPC ingest port. |
| otlp_http_port | number | 4318 | OTLP HTTP ingest port. |
| block_retention | string | "336h" | How long to keep trace blocks before compaction removes them. |
| data_volume | string | "tempo_data" | Named volume mounted at /var/tempo (WAL + trace blocks). |
| constraints | list | [] | Placement constraints. On a nomploy cluster: attribute = "$${meta.nomploy_control_plane}", operator = "=", value = "true". |
| resources | object | {
cpu = 300
memory = 256
} | The task resources. |
No variables match.
This pack stores data in one Docker named volume:
tempo_data
restic
# Run on the node hosting this pack. Point restic at your repo first: # export RESTIC_REPOSITORY="s3:https://<account>.r2.cloudflarestorage.com/<bucket>" # export RESTIC_PASSWORD="<repo-password>" # export AWS_ACCESS_KEY_ID=<key> AWS_SECRET_ACCESS_KEY=<secret> restic backup \ /var/lib/docker/volumes/tempo_data/_data
rclone (sync to S3/R2)
rclone sync /var/lib/docker/volumes/tempo_data/_data backup:<bucket>/tempo_data
Paths assume the default Docker volume location (/var/lib/docker/volumes). Restore by stopping the job, restoring files into the same volume, and re-running the pack.
Grafana Tempo — a high-scale, cost-efficient distributed tracing backend. It ingests spans over OTLP and stores trace blocks on local disk (or object storage), and you explore traces from Grafana. Together with Loki (logs), Prometheus (metrics) and Pyroscope (profiles) it completes the Grafana observability stack.
Single host-networked Nomad service. This pack renders a minimal single-binary config and stores data on a volume.
nomad-pack registry add nomploy https://github.com/Nomploy/nomad-packs
nomad-pack run tempo --registry=nomploy
| Variable | Default | Description |
|---|---|---|
port |
3200 |
HTTP API / query port (server.http_listen_port). |
otlp_grpc_port |
4317 |
OTLP gRPC ingest port. |
otlp_http_port |
4318 |
OTLP HTTP ingest port. |
block_retention |
336h |
How long to keep trace blocks (14 days). |
image |
grafana/tempo:latest |
Container image. Pin a tag in production. |
data_volume |
tempo_data |
/var/tempo — WAL and trace blocks. |
resources |
{ cpu = 300, memory = 256 } |
Task resources. Bump for heavy ingestion. |
Send spans from your apps or an OTel Collector to
<host>:4317(gRPC) or<host>:4318(HTTP). Add Tempo as a data source in grafana (URLhttp://<host>:3200) to explore traces. A prestart init task chowns the data volume to uid10001. The bundled config uses local block storage; point it at rustfs/S3 for larger deployments. Pin the job withconstraints.