What performance overhead does SigNoz add? Is it s...
# support
s
What performance overhead does SigNoz add? Is it something that is suitable for production environments or only lower development environments?
v
SigNoz supports open-telemetry instrumentation which allows generating telemetry data with very less performance overhead. Open-telemetry is industry standard created by companies like Google, Uber, CNCF, etc. Major companies already uses it. More details here: https://signoz.io/blog/opentelemetry-tracing/ SigNoz is suitable for production environments. Here are few tips for running SigNoz on production: https://signoz.io/docs/production-readiness/
s
Since the JVM is started with an agent, are there risks associated like crashing to process due to an issue?
a
we currently run the OTEL agents in PROD with traffic into the millions, no issue so far.
not a stability issue directly, but some services do get their logs flooded with these weird warnings which i haven’t pinpointed on how to solve:
Copy code
[otel.javaagent 2023-02-16 22:31:53:598 +0000] [PeriodicMetricReader-1] WARN io.opentelemetry.sdk.metrics.internal.state.AsynchronousMetricStorage - Instrument db.client.connections.pending_requests has recorded multiple values for the same attributes.
s
@Alexei Zenin good to know, thanks for sharing