Help! Something is wrong with my Jaeger installation!

Help! Something is wrong with my Jaeger installation!

  • January 23, 2019
Table of Contents

Help! Something is wrong with my Jaeger installation!

Jaeger is composed of different components, each potentially running in its own host. It might be the case that one of these moving parts isn’t working properly, causing spans to not be processed and stored. When something goes wrong, make sure to check the items listed here.

Before everything else, make sure to confirm what is the sampling strategy. By default, Jaeger uses a probabilistic sampling strategy, with a 1/1000th chance that the span will be reported. Typically, the sampling strategy can be set via the environment variables JAEGER_SAMPLER_TYPEand JAEGER_SAMPLER_PARAM, but refer to the Jaeger Client’s documentation for the language you are using for more details about which sampling strategies are available.

When using the Jaeger Java Client, the strategy is usually printed out via the logging facility provided by the instrumented application when creating the tracer: Some Jaeger clients are able to log the spans that are being reported to the logging facility provided by the instrumented application. Typically, this can be done by setting the environment variable JAEGER_REPORTER_LOG_SPANSto true, but refer to the Jaeger Client’s documentation for the language you are using. When using the Jaeger Java Client, spans are reported like the following:

Source: medium.com

Share :
comments powered by Disqus

Related Posts

Deployment strategies for the Jaeger Agent

Deployment strategies for the Jaeger Agent

If you’ve been following the evolution of the Kubernetes templates for Jaeger, you might have noticed an important change recently: the Jaeger Agent is not being deployed as a DaemonSet anymore. Instead, instructions are now being provided on how to deploy it as a “Sidecar”. The Agent component was developed to act as a “buffer” between the tracer and the collector.

Read More
Rate Limiting at the Edge

Rate Limiting at the Edge

I’m sure many of you have heard of the “Death Star Security” model—the hardening of the perimeter, without much attention paid to the inner core—and while this is generally considered bad form in the current cloud native landscape, there is still many things that do need to be implemented at edge in order to provide both operational and business logic support. One of these things is rate limiting. Modern applications and APIs can experience a burst of traffic over a short time period, for both good and bad reasons, but this needs to be managed well if your business model relies upon the successful completion of requests by paying customers.

Read More