Obzly
Compare / Grafana LGTM

Obzly vs Grafana LGTM

Grafana's stack is the most common way to self-host observability: Grafana for dashboards, Loki for logs, Mimir for metrics and Tempo for traces. Obzly covers the same three signals in one system, and it can also sit behind the Grafana you already have.

Checked 27 October 2026 against Grafana LGTM's own docs and pricing. Sources

Which to choose

Choose Obzly if

  • you'd rather run one system than four, each with its own components and tuning;
  • you want SSO, SCIM, fine-grained roles and an audit log without an enterprise license;
  • you want synthetic checks and an infrastructure inventory that you host yourself;
  • you need one login across regions, with each region's data staying put;
  • Apache 2.0 suits your legal team better than AGPLv3.

Choose Grafana LGTM if

  • you run metrics at very large scale: Mimir is built for over a billion active series;
  • you need browser synthetic checks, real-user monitoring or continuous profiling;
  • you visualise many data sources that Obzly doesn't store, such as CloudWatch or SQL databases;
  • you want a managed service with a free tier (Grafana Cloud);
  • you need SAML, which Obzly doesn't support.

What you run

To self-host Grafana LGTM you run Grafana, Loki, Mimir, Tempo, with object storage. Obzly is one image in a few roles, plus Postgres, Valkey and object storage.Grafana LGTM, self-hostedGrafanaUI · MySQL or PostgreSQL for HALokidistributoringesterqueriercompactorruler · …Mimirdistributoringesterstore-gatewaycompactorquery-frontendTempodistributoringesterqueriercompactorObject storage · configured per backendObzlyone imageSERVICE_MODE = all, or one role per deployment:apiworkerbrokerquerierlogs, metrics and traces in one storeplus control-plane for many SitesPostgresValkeyObject storage
Loki and Mimir can also run as a single binary for small setups; at scale each splits into the components shown. Obzly scales the same way, one role at a time, from the same image.

Feature by feature

CapabilityObzlyGrafana OSS stackGrafana Enterprise / CloudSource
LicenseApache 2.0, every featureAGPLv3 (Grafana, Loki, Tempo, Mimir)ProprietaryLicensing
What you runOne image in up to five roles, Postgres, ValkeyGrafana plus Loki, Mimir and Tempo, each a set of components (or single-binary modes)Same, or nothing with Grafana CloudLoki componentsMimir components
StorageZSTD Parquet on S3, GCS, Azure Blob or MinIOObject storage for logs, metrics and tracesSame; Cloud stores it for youLoki storageTempo storage
OTLP inLogs, metrics and traces, gRPC and HTTPAll three; Loki and Mimir over HTTPAll threeLoki OTLPMimir OTLP
Query languagesSearch bar and SQL for logs, PromQL, Jaeger API for tracesLogQL, PromQL, TraceQLSameTraceQL
PrometheusRemote-write ingest and a PromQL APIRemote write and PromQL, nativelySameMimir
DashboardsOwn dashboards; imports Grafana layout without queriesNative, with many data sourcesSameImport dashboards
AlertingWebhook, Slack, PagerDuty, email20+ contact pointsSameContact points
Synthetic checksHTTP, TLS, TCP, DNS, ICMP from your probes; no browserNot availableGrafana Cloud, incl. browser and private probesSynthetic Monitoring
Infrastructure inventoryClusters to containers, plus AWS and Azure resourcesNo inventory viewEntity catalog of services, pods and nodes (Cloud)Entity catalogMonitor infrastructure
RUM and profilingNeitherPyroscope for profilingFrontend Observability (Cloud), PyroscopeFrontend ObservabilityPyroscope
SSOOIDC with any provider; no SAMLOAuth and LDAP; no SAMLSAML in Enterprise and Cloud Pro and upSAMLDeployment options
SCIMIncludedNot availableEnterprise and CloudSCIM
RBAC and audit logPer-project roles, teams, audit log with CSV exportBasic roles onlyFine-grained RBAC and audit logRBACAudit log
MCP serverBuilt in at /mcp, read-onlyOpen-source server you runPlus a hosted server for CloudGrafana MCP
Many regionsControl plane in front of many SitesOne Grafana over many data sourcesCross-cluster query federation in GEM, experimentalGEM federation
PriceFree; you run the infrastructureFree; you run the infrastructureCloud Pro from $19/month plus usage; Enterprise from $25k/yearPricing

Use both

You don't have to give up Grafana. Obzly speaks the APIs Grafana already uses, so it can replace Loki, Mimir and Tempo underneath while your team keeps its dashboards.

Prometheus data source

Point it at Obzly's PromQL API. PromQL panels and alert queries keep working where metric and label names match; OTLP attributes are stored without dots, for example k8s_pod_name.

Jaeger data source

Point it at Obzly's Jaeger-compatible API to browse traces in Grafana.

Logs

LogQL panels need rewriting. Use Obzly's log search or SQL for those instead.

Moving from Grafana LGTM

  1. Send to both. Add an OTLP exporter for Obzly to your Collector or Alloy pipeline next to the existing one. Collector guide
  2. Keep Prometheus flowing. Add Obzly as a remote-write target. PromQL and Grafana
  3. Repoint Grafana. Add Obzly as a Prometheus and a Jaeger data source and switch panels over.
  4. Bring dashboards into Obzly, if you want them there. Importing a Grafana dashboard copies its panels and layout; you add each panel's query in Obzly.
  5. Move alert rules. PromQL rules carry over once metric and label names are checked; log rules need Obzly's search syntax. Alerts
  6. Switch off the old backends once Obzly has your full retention.

Sources

Every Grafana LGTM fact on this page comes from its vendor's own pages, read on 27 October 2026. Plan names and prices change; check the linked page before you decide.

Spotted something out of date? Open an issue.