Closed trentm closed 2 years ago
The first commit here is a first stab that has the APM Server fetching retry if it fails on earlier attempts. IIUC the java agent does not retry (see https://github.com/elastic/apm-agent-java/blob/master/apm-agent-core/src/main/java/co/elastic/apm/agent/report/ApmServerHealthChecker.java). I think I am going to remove the retrying in the node.js agent, but I am not sure yet. If there are no retries, then if the APM Server happens to be down when the agent starts, then the agent will stay in a "degraded" state not knowing the APM server version. If this "degraded" state just assumes the latest APM Server version then that is perhaps fine, but it could mean not getting data to the (older) APM server version if there is a compat issue.
the below badges are clickable and redirect to their specific view in the CI or DOCS
#### Build stats * Reason: `null` * Start Time: 2022-01-25T23:33:24.385+0000 * Duration: 11 min 39 sec * Commit: fca8cf1063a481c8cb07caadb23dc9cf8b814ef9 #### Test stats :test_tube: | Test | Results | | ------------ | :-----------------------------: | | Failed | 0 | | Passed | 17206 | | Skipped | 0 | | Total | 17206 |
To re-run your PR in the CI, just comment with:
run
elasticsearch-ci/docs
: Re-trigger the docs validation. (use unformatted text in the comment!)elastic-apm-http-client@10.4.0 published v10.4.0 tagged
This is to support the Node.js APM agent changing its behaviour based on the APM Server version. Specifically whether to send or drop unsampled transactions.
Refs: https://github.com/elastic/apm-agent-nodejs/issues/2455