[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC):2025-04-03。"],[[["This document outlines the steps to capture a memory heap snapshot of the AlloyDB Omni operator for diagnosing memory-related issues."],["The process involves identifying one of two AlloyDB Omni operator deployments (`fleet-controller-manager` or `local-controller-manager`) to analyze."],["Memory analysis is enabled by adding the `--pprof-address` argument with an available port to the chosen deployment's configuration, which triggers a pod restart."],["After enabling memory analysis, a memory heap snapshot can be created by using `curl` command that saves it to a file named `heap.out`, accessible via port forwarding."],["After analysis, it's critical to disable memory analysis by removing the `pprof-address` argument from the deployment configuration to prevent potential security risks."]]],[]]