Skip to content

Logchef vs Datadog Log Management

Logchef is a self-hosted, open source alternative to one part of Datadog: log search, log dashboards, and log alerting. It is not a Datadog replacement. Datadog is a SaaS observability platform that collects, processes, and stores your logs and ties them to traces, metrics, and security products. Logchef is a query UI over a ClickHouse, VictoriaLogs, or Grafana Loki backend that you run yourself.

Pick Datadog if you want one vendor to collect, parse, store, and correlate logs with APM and infrastructure monitoring, and you do not want to operate a log database. That is a real product advantage, and Logchef does not match it.

Pick Logchef if you already run ClickHouse, VictoriaLogs, or Loki (or are willing to), you want every stored log to be searchable, and you want logs, queries, and access rules to stay on your own infrastructure. You take on the work Datadog does for you: shipping logs, sizing the database, and setting retention.

You can keep Datadog for APM and infrastructure and move only logs. See Using them together.

This page was checked on 2026-11-18 against Datadog’s public docs and pricing page, and the Logchef repository.

Logchef Datadog Log Management
What it is A log search, dashboard, and alerting UI over your own ClickHouse, VictoriaLogs, or Loki The log product of a SaaS observability platform
License AGPLv3 Proprietary service. The Datadog Agent is Apache 2.0
Storage Your ClickHouse tables, VictoriaLogs instances, or Loki tenants Datadog-hosted Standard and Flex tiers, plus archives in your own S3, Azure Storage, or GCS bucket
Ingestion None. You run a collector such as Vector or the OpenTelemetry Collector Included: collection from hosts, containers, and cloud providers, then pipelines and processors
Query languages LogchefQL, ClickHouse SQL, LogsQL, LogQL Datadog’s log search syntax
Signals Logs Logs connected to traces and metrics, and Cloud SIEM on the same logs
Deployment One binary plus your log backend SaaS, with Agents on your hosts
Access control OIDC or local login. Teams scope which sources a user can query. An audit log for admins Roles and permissions. A restriction query limits which logs a role can read
Alerting Scheduled rules in LogchefQL, SQL, LogsQL, or LogQL, with email, webhook, Slack, and PagerDuty destinations Log monitors over indexed logs, with notifications to Slack, Jira, webhooks, Microsoft Teams, PagerDuty, and email
Dashboards Log panels over your log sources, with scheduled email reports Datadog dashboards across the platform’s products
AI and MCP An AI query assistant that uses your own model provider, and a built-in read-only MCP endpoint A remote Datadog MCP Server with a search_datadog_logs tool
Pricing No license fee. You pay for the infrastructure you run Per GB ingested, plus per million events indexed by retention. See Pricing

Datadog Log Management covers the whole path. Per its docs, it ingests logs from hosts, containers, and cloud providers, enriches them with pipelines and processors, stores them, and connects them to traces and metrics.

Logchef covers the last step only. It does not ingest logs. A collector writes logs into ClickHouse, VictoriaLogs, or Loki, and Logchef reads them. You own the collector configuration, the table schema, the retention setting, and the database upgrades. You can keep the logs, saved queries, and UI inside your network. Optional external AI providers and notification destinations receive the data you send to them.

Datadog lists ingestion, Standard indexing, and Flex pricing on its pricing page on 2026-11-18:

  • Ingest starts at $0.10 per GB of uncompressed data per month, billed annually or on demand. This covers processing, live tail, and forwarding to your own archive bucket.
  • Standard indexing costs $1.70 per million log events per month at 15-day retention when billed annually, or $2.55 on demand.
  • Flex Logs is metered per million events stored, with each million billed for at least 30 days at a minimum of $0.05. Flex compute is sized separately.
  • Rehydrating archived logs costs $0.10 per compressed GB scanned, plus indexing for the events it brings back.
  • Forwarding logs to other destinations, such as an external SIEM, starts at $0.25 per uncompressed GB per destination.

Check the pricing page for current numbers before you compare.

Logchef has no license fee, no per-GB fee, and no per-user fee. Your cost is the backend infrastructure and the engineering time to run it. This page makes no claim about which is cheaper for you. That depends on your volume, your retention, and the cost of that time. The structural difference is the unit: Datadog charges per event and per GB, and a self-hosted backend costs what its disks and CPUs cost.

Datadog separates ingestion from indexing and calls this Logging without Limits. You ingest everything and index a subset. Logs that an exclusion filter drops from an index remain available in Live Tail and can feed log-based metrics or configured archives. They are excluded from indexed search. Three limits follow:

With Logchef there are no tiers in the UI. Searches and alert rules query the configured source, subject to access controls, time ranges, and query limits. Retention is the TTL of your ClickHouse table or the retention setting of VictoriaLogs or Loki. The trade-off is direct. Datadog lets you cut cost by indexing less. With your own backend you cut cost by storing less or by adding cheaper disks.

Datadog’s search syntax uses @attribute:value terms, tags, and boolean operators. Numeric range queries need the attribute to be a facet first.

Logchef has two levels. LogchefQL is a short filter syntax, such as status>=500 and service="checkout", with aggregation pipes for counts and percentiles. Native mode accepts a single ClickHouse SELECT, including joins, CTEs, and window functions. Logchef parses the SQL, expands time macros, and applies result limits.

Datadog’s log product is tied to the rest of the platform: traces, metrics, and Cloud SIEM, which runs on the same ingested logs. Logchef has no traces UI, no APM, no metrics store, and no session replay. A ClickHouse table can keep a trace_id column and you can filter on it, but Logchef will not open the trace.

You can keep Datadog for APM and infrastructure and send logs to both places.

  • Vector’s datadog_agent source receives logs from a Datadog Agent over HTTP. Add a ClickHouse or VictoriaLogs sink as described in Shipping Logs with Vector, and your hosts keep the Agent they already run.
  • Index a small, high-value subset in Datadog for correlation with traces. Keep the full stream in ClickHouse and search it with Logchef.
  • Datadog archives are compressed JSON files in a bucket you own, so your log history is not locked in.
  • You want logs, traces, and infrastructure metrics correlated in one product. Logchef does logs only.
  • You do not want to operate a log backend or a collector fleet.
  • You rely on Datadog’s integrations and processing pipelines to parse logs from many sources.
  • You need Cloud SIEM or other security products on the same log stream.
  • You need alert notifications to Jira or Microsoft Teams as built-in targets. Logchef has email, webhook, Slack, and PagerDuty destinations.
  • You need log data to stay in your cloud account but want Datadog’s UI. BYOC Logs stores logs in your object storage and serves queries from the Datadog UI.
  • You already run ClickHouse, VictoriaLogs, or Loki, or your team can.
  • You want to search and alert on every stored log, without index and exclusion decisions.
  • The whole stack must be self-hosted, UI and metadata included. Logchef is one binary with a SQLite or Postgres metadata store.
  • You want SQL over your logs.
  • You want open source software, with team-scoped access, a CLI, and an MCP endpoint in the same release.

There is no Datadog-specific migration guide yet. The path has three steps:

  1. Pick a backend and create a table. Start from the schema design guide. Datadog’s reserved attributes, such as service and status, map to columns.
  2. Ship logs to the backend with Vector or the OpenTelemetry Collector.
  3. Install Logchef with the quickstart and add the table as a source.

Simple filters translate directly. Datadog’s service:checkout status:error becomes service="checkout" and status="error" in LogchefQL. A range such as @http.status_code:[500 TO 599] becomes http.status_code>=500 and http.status_code<=599, because LogchefQL reads nested keys with dot notation.

Is Logchef an open source Datadog alternative?

Section titled “Is Logchef an open source Datadog alternative?”

For log search, log dashboards, and log alerting, yes. For APM, infrastructure monitoring, or security products, no. Logchef also needs a ClickHouse, VictoriaLogs, or Loki backend that you run.

Not the whole product. Datadog’s BYOC Logs runs log storage in your own infrastructure, and the UI stays in Datadog’s SaaS. Logchef and its backends run entirely on your side.

No. Logchef does not collect logs. Use Vector, the OpenTelemetry Collector, or keep the Datadog Agent and route its logs through Vector.

Can I keep Datadog APM and move only logs?

Section titled “Can I keep Datadog APM and move only logs?”

Yes. You lose the in-app jump between a trace and its logs. Keep the trace ID as a column so you can still filter logs by it in Logchef.

In two parts: a per-GB ingest fee and a per-million-events indexing fee that depends on retention. Flex Logs adds storage and compute fees for longer retention. Logchef has no usage-based fee.

Does Logchef have log tiers like Flex Logs or archives?

Section titled “Does Logchef have log tiers like Flex Logs or archives?”

No. Logchef reads what the backend holds. Retention and storage tiering are settings of that backend, not of Logchef.

To try it, start with the quickstart.