What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
To target 7% of Langfuse traces, set the client-side sampling rate to 0.07. That intentionally gives you partial observability: Langfuse makes the decision per trace, and an unsampled trace sends none of its observations or associated scores. Separately, ClickHouse system logs can keep using disk even after Langfuse trace retention is configured. The two retention mechanisms need to be checked independently.
How do I sample only 7% of Langfuse traces?
Configure sampling in the Langfuse client or SDK. The documented setting accepts a value from 0 to 1; 0.07 is a target sampling rate of 7%. Langfuse documents a default of 1, meaning all traces are collected. See the Langfuse sampling documentation for the configuration supported by your SDK.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Up and Running with ClickHouse: Learn and Explore ClickHouse, It's Robust Table Engines for... | $19.95 | Buy on Amazon |
The equivalent setting is named sample_rate or sampleRate depending on the client, or can be supplied through the LANGFUSE_SAMPLE_RATE environment variable. This is a target rate, not proof that exactly 7% of a particular agent’s executions were captured. The title’s 7% figure is not independently verified by execution data.
What the rate leaves out
Sampling is decided at trace level, not independently for each observation. When a trace is not sampled, its observations and associated scores are not sent. A 7% setting therefore reduces trace volume at the cost of leaving most runs without Langfuse trace data. Do not treat sampled traces as a complete activity ledger or use their absence to conclude that an agent did not run.
Why is ClickHouse logging itself?
ClickHouse is currently a required component of self-hosted Langfuse; Langfuse’s self-hosting documentation says no alternative OLAP database is supported for this deployment. ClickHouse also maintains its own system-log tables for server and diagnostic activity. These are separate from the Langfuse tracing data stored and managed by the application.
Langfuse’s ClickHouse self-hosting guide warns that system tables such as trace_log, text_log, and opentelemetry_span_log have no TTL by default and can contribute to disk growth. “Logging itself” is therefore normal server behavior; whether it explains a particular installation’s disk use depends on which tables are growing and how that ClickHouse instance is configured.
Why can disk usage grow after I set Langfuse retention?
Langfuse trace retention and ClickHouse system-log retention govern different data. Configuring retention for tracing data or media does not, by itself, establish a retention policy for ClickHouse’s server system logs. Langfuse specifically identifies system-log tables as a common explanation when ClickHouse disk use continues to rise despite retention being configured.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Diagnose the installation before changing retention
- Identify which ClickHouse tables are using disk and whether the growth is in Langfuse data or system-log tables such as
system.trace_log,system.text_log, orsystem.opentelemetry_span_log. - Check the Langfuse release and the ingestion path in use; behavior and table layouts can vary by version.
- Review ClickHouse’s configuration and the TTLs actually applied to the relevant system tables. Langfuse documents a TTL example for
system.trace_log, but the appropriate retention period depends on how much diagnostic history the operator needs. - Set and verify system-log retention deliberately rather than assuming the Langfuse trace-retention setting covers it. Shorter diagnostic retention can limit disk use but also reduces the history available for troubleshooting.
The title alone does not establish that ClickHouse system logs are causing a specific disk problem. A diagnosis requires the growing table names, Langfuse version, ingestion path, and relevant ClickHouse settings.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Check ingestion compatibility before treating missing traces as a sampling problem
Langfuse documents OpenTelemetry as its supported trace-ingestion path. Its OpenTelemetry documentation and v4 compatibility guidance describe possible ingestion delays when older SDKs or exporters omit the v4 ingestion header. Check the installed SDK or exporter and its compatibility with the deployed Langfuse release when traces are delayed or missing; a sampling setting is not the only possible explanation.
Version-specific v4 storage terms
In Langfuse v4, events_full is the immutable, full-fidelity event table, while events_core is a query-optimized projection. LLM calls, tool executions, and agent steps are represented as rows with trace-level attributes. These names and this data model are specific to v4; do not assume they describe older installations.
Keep sensitive data out before it leaves the application
Langfuse documents client-side masking for sensitive content. Centralized server-side ingestion masking for self-hosted deployments is an Enterprise feature, and Langfuse describes it as a safety net rather than a replacement for client-side masking when data must not leave the application boundary. See the Langfuse data-masking documentation for the applicable controls.
Free tools Windows power users keep installed
One-click scans. No signup required.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
