IMPORTANT: No additional bug fixes or documentation updates will be released for this version. For the latest information, see the current release documentation.
« Log correlation Release notes »
Elastic Docs ›

Troubleshooting

edit

Troubleshooting

edit

Hopefully the agent Just Works™, but depending on your situation the agent might need some tuning.

First, to learn more about what’s going on inside the agent, you can increase the amount of log messages it writes. To do this, set the log level with the option log_level = 0 — 0 being the level of most messages, DEBUG.

In your config/elastic_apm.yml:

log_level: <%= Logger::DEBUG %>

Log messages

edit

Queue is full (256 items), skipping…

edit

The agent has an internal queue that holds events after they are done, and before they are safely serialized and sent to APM Server. To avoid using up all of your memory, this queue has a fixed size. Depending on your load and server setup, events may be added to the queue faster than they are consumed, hence the warning.

Things to consider:

  • Is server_url misconfigured or APM Server down? If the agent fails to connect you will also see log messages containing Connection error or Couldn't establish connection to APM Server.
  • Experiencing high load? The agent can spawn multiple instances of its Workers that pick off the queue by changing the option pool_size (default is 1).
  • If you have high load you may also consider setting transaction_sample_rate to something smaller than 1.0. This determines whether to include spans for every transaction. If you have enough traffic, skipping some (probably) identical spans won’t have a noticeable effect on your data.
« Log correlation Release notes »