gke-ai-troubleshooting-tpu-connection-failure-vbar-oom

TPU Connection Failure and VBAR OOM Troubleshooting

Use this skill to systematically diagnose and prevent vbar_control_agent segfaults and Out-Of-Memory (OOM) errors on TPU v6e nodes.

⚠️ Prerequisites

• Cloud Logging must be enabled for the project.
• Access to the project and cluster via gcloud or equivalent tool.

🔍 Diagnostic Workflow

Step 0: Context Acquisition & Time Window Definition

To begin troubleshooting, acquire the following context from the user:

• Project ID (e.g., customer-ai-project-123)
• Cluster Name (e.g., tpu-cluster-prod)
• Node Name or Instance ID (e.g., tpu-node-1)
• Workload Name (JobSet Name) (e.g., my-training-job-456)
• Workload Namespace
• Issue Time (e.g., 2026-04-14T20:00:00Z)

Time Handling Rules

1. Reject Relative Time: If the user says "X minutes ago" or "just now", stop and ask for the exact timestamp or a specific time window.
2. Window Calculation: If the user provides a start time or an "around" time T, calculate the query window as [T - 30m] to [T + 30m].
   • Let Start_Time = T - 30m
   • Let End_Time = T + 30m

Step 1: Check for vbar_control_agent OOMs

Look for specific out of memory messages from vbar_control_agent in serial console logs.

• Tool to use: query_logs
• Filter Templates:

Serial Console Logs (OOMs):

logName="projects/<project_id>/logs/serialconsole.googleapis.com%2fserial_port_1_output"
AND labels."compute.googleapis.com/resource_name"="<node_name>"
AND SEARCH(text_payload, "Memory cgroup out of memory: Killed process .* (vbar_control_ag)")
AND timestamp >= "<Start_Time>"
AND timestamp <= "<End_Time>"

• Logic: Presence of Memory cgroup out of memory messages related to vbar_control_agent. Stack traces pointing to libtpu::tpunetd::VBARControlHelper::MetricsReadFromVBAR are a strong indicator.
• Automation: Proceed to next step automatically after reporting findings.
• Reference: See references/failure_signatures.md for example log patterns.

Step 2: Investigate tpu-device-plugin Metrics Fetch Failures [Low Risk]

Check if tpu-device-plugin is reporting metric fetch failures.

• Tool to use: query_logs
• Filter Template:

resource.type="k8s_container"
AND resource.labels.project_id="<project_id>"
AND resource.labels.cluster_name="<cluster_name>"
AND resource.labels.container_name="tpu-device-plugin"
AND severity=ERROR
AND textPayload:"metrics fetch failed for .* deviceID and .* device path with error: checksum didn't match with the metrics data. Corrupt data found"
AND timestamp >= "<Start_Time>"
AND timestamp <= "<End_Time>"

• Logic: Errors indicating "metrics fetch failed" with "checksum didn't match" suggest vBAR memory corruption.
• Automation: Proceed to next step automatically after reporting findings.

Step 3: Check for Custom Metrics Collection Usage [Low Risk]

Inquire with the user about any custom TPU metrics collection mechanisms they have deployed.

• Action: Ask the user if they are using custom scripts or agents (e.g., using libtpu.sdk.tpumonitoring) that frequently query GetHostMetrics from vBAR Control Agent.
• Logic: Confirmation of custom metrics collection helps confirm the race condition hypothesis.
• Automation: Stop and wait for user response before proceeding to resolution.

🛠️ Resolution Workflow

Resolution 1: Temporarily Disable Custom Metrics Collection [High Risk]

If a custom metrics collection agent is identified, advise the user to temporarily disable it.

• Action: Recommend disabling the custom metrics collector.
• Justification: Prevents reads from vBAR during device resets, stopping crashes and OOMs.
• Automation: Stop and request explicit user approval in the bug thread before making this recommendation or taking action.

Resolution 2: Await vbar_control_agent Resiliency Update [Low Risk]

Advise the user that a permanent fix will be available in a future GKE version.

• Action: Recommend upgrading GKE when the fix is available.
• Justification: The updated agent will be resilient to memory corruption and gracefully handle reads from unbound vBARs.
• Automation: Proceed to report this finding.

📋 copypaste checklist

• Acquire context and compute [T - 30m, T + 30m] window.
• Check for vbar_control_agent segfaults and OOMs using query_logs.
• Investigate tpu-device-plugin failures using query_logs.
• Ask user about custom metrics collection usage.
• Advise disabling custom metrics collection (High Risk) if applicable.
• Advise awaiting resiliency update.