Stay organized with collections
Save and categorize content based on your preferences.
This page describes known issues that you might run into while running your
AI-optimized VMs or clusters. For issues with Compute Engine VMs, see
Compute Engine known issues.
Issues
The following section list known issues and workarounds for AI Hypercomputer.
The metadata server might display old physicalHost VM metadata
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-09-04 UTC."],[[["\u003cp\u003eThis page outlines potential issues encountered while using AI Hypercomputer, differentiating them from Compute Engine VM issues, which have their own documentation.\u003c/p\u003e\n"],["\u003cp\u003eAI Hypercomputer is currently in a pre-GA stage, meaning it's provided "as is" with potentially limited support, subject to the Pre-GA Offerings Terms.\u003c/p\u003e\n"],["\u003cp\u003eAfter a host error or using the report faulty host API, the metadata server might display outdated \u003ccode\u003ephysicalHost\u003c/code\u003e information for an instance.\u003c/p\u003e\n"],["\u003cp\u003eTo get correct \u003ccode\u003ephysicalHost\u003c/code\u003e data, you can utilize the \u003ccode\u003einstances.get\u003c/code\u003e method, \u003ccode\u003egcloud compute instances describe\u003c/code\u003e command, stop and restart your instance, or wait up to 24 hours for automatic updates.\u003c/p\u003e\n"]]],[],null,["This page describes known issues that you might run into while running your\nAI-optimized VMs or clusters. For issues with Compute Engine VMs, see\n[Compute Engine known issues](/compute/docs/troubleshooting/known-issues).\n\nIssues\n\nThe following section list known issues and workarounds for AI Hypercomputer.\n\nThe metadata server might display old `physicalHost` VM metadata\n\nAfter experiencing a\n[host error](/compute/docs/instances/host-maintenance-overview#hosterror) or\nusing the\n[report faulty host API](/ai-hypercomputer/docs/manage/report-faulty-host) to\nmove an instance to a new host, when you\n[query the metadata server](/compute/docs/metadata/querying-metadata#programmatically-query-metadata),\nit might display the\n`physicalHost` metadata of the instance's previous host.\n\nTo workaround this issue, do one of the following:\n\n- Use the [`instances.get` method](/compute/docs/reference/rest/v1/instances/get) or the [`gcloud compute instances describe` command](/sdk/gcloud/reference/compute/instances/describe) to retrieve the correct `physicalHost` information.\n- [Stop and then start](/compute/docs/instances/stop-start-instance) your instance. This process updates the `physicalHost` information in the metadata server.\n- Wait 24 hours for the impacted instance's `physicalHost` information to be updated."]]