EMR on ACK runs Presto jobs inside containers that are deleted after each job completes. This means runtime logs are lost unless you persist them before the container is removed. Enable Simple Log Service (SLS) to collect and store container logs so you can diagnose failures after the fact.
Prerequisites
Before you begin, ensure that you have:
A Presto cluster created on the EMR on ACK page of the new EMR console. See Get started with EMR on ACK
Simple Log Service activated. See Getting Started
Collect Presto container logs
Use DaemonSet mode to collect container logs from your Presto cluster. For the full setup procedure, see Collect text logs from Kubernetes containers in DaemonSet mode.
When configuring log paths in the DaemonSet setup, use the following paths based on what you need to capture:
| Log type | Log path | What it captures |
|---|---|---|
| Standard output | stdout | Job progress, query results, and general runtime output |
| Core dump | /data/trino/hs_err_pid1.log | JVM crash details and fatal error information |