hello, I have a python application (flask) connected with sentry and the traces on new relic show a large amount of time in the sentry package(raven-python). Is this a sentry issue or new relic issue with dealing with my app stack?
Its hard to say what its measuring. Network requests are default async in Sentry, and thats going to be the majority of time consumption. We do constantly collect data via breadcrumbs, the performance impact should be negligible.
Hello zeeg, not sure what you mean with “collect data via breadcrumbs”. Can you explain to me?
Can you please take a look on the trace of some route form me? as you can see segments of high time consumption on Application code (in raven.middleware:Sentry), do you think is this a issue on your library, on the way newrelic parses the data or it is a configuration problem?
You can try disabling breadcrumb collection (see our docs for configuration options).
Breadcrumbs are a feature of Sentry which capture trailing events (like log entries). It’s the only active instrumentation Sentry has today, and the only thing that could realistically create a performance hit unless you’re already in an exception path.