vastai-incident-runbook

Vast.ai Incident Runbook

Overview

Rapid incident response procedures for Vast.ai GPU instance failures. Covers triage, mitigation, recovery, and postmortem for common incident types: spot preemption, instance crashes, GPU failures, and billing issues.

Prerequisites

• Vast.ai CLI access
• SSH access to instances (if still running)
• Checkpoint storage accessible (S3/GCS)

Instructions

Triage: Assess Impact (< 2 minutes)

#!/bin/bash
set -euo pipefail
echo "=== INCIDENT TRIAGE ==="
echo "Time: $(date -u)"

# 1. Check all instances
echo -e "\n--- Instance Status ---"
vastai show instances --raw | python3 -c "
import sys, json
for inst in json.load(sys.stdin):
    status = inst.get('actual_status', '?')
    flag = 'ALERT' if status in ('error', 'exited', 'offline') else 'OK'
    print(f'  [{flag}] ID:{inst[\"id\"]} Status:{status} '
          f'GPU:{inst.get(\"gpu_name\",\"?\")} \${inst.get(\"dph_total\",0):.3f}/hr')
"

# 2. Check if affected instance has recent logs
echo -e "\n--- Recent Logs (last 20 lines) ---"
vastai logs ${INSTANCE_ID:-0} --tail 20 2>/dev/null || echo "No logs available"

# 3. Check account balance
echo -e "\n--- Account ---"
vastai show user --raw | python3 -c "import sys,json; u=json.load(sys.stdin); print(f'Balance: \${u.get(\"balance\",0):.2f}')"

Incident Type 1: Spot Preemption

Symptoms: Instance status changes from running to exited or offline without user action.

# 1. Verify preemption (not user error)
vastai show instance $ID --raw | python3 -c "
import sys, json; i=json.load(sys.stdin)
print(f'Status: {i.get(\"actual_status\")}')
print(f'Status msg: {i.get(\"status_msg\", \"none\")}')
"

# 2. Check if checkpoint was saved
# (depends on your checkpoint storage — S3, GCS, etc.)
aws s3 ls s3://bucket/checkpoints/ --recursive | tail -5

# 3. Provision replacement instance
vastai search offers "gpu_name=${GPU_NAME} reliability>0.98 rentable=true" \
  --order dph_total --limit 3

# 4. Create replacement and resume from checkpoint
vastai create instance $NEW_OFFER_ID --image $IMAGE --disk 50

Incident Type 2: Training Job Crash

Symptoms: Instance running but training process exited with error.

# 1. SSH in and check logs
ssh -p $PORT root@$HOST "tail -100 /workspace/train.log 2>/dev/null || echo 'No log file'"

# 2. Common causes
ssh -p $PORT root@$HOST << 'CHECK'
# GPU memory issue?
nvidia-smi | grep -i "out of memory" && echo "OOM detected"
# Disk full?
df -h /workspace | tail -1
# Process still running?
ps aux | grep python | grep -v grep
CHECK

# 3. Restart training from checkpoint
ssh -p $PORT root@$HOST "cd /workspace && python train.py --resume-from latest"

Incident Type 3: GPU Hardware Failure

Symptoms: nvidia-smi fails, CUDA errors, or ECC memory errors.

# 1. Check GPU health
ssh -p $PORT root@$HOST "nvidia-smi" || echo "GPU not responding"

# 2. This is a host-level failure — you cannot fix it
# Destroy the instance and provision on a different host
vastai destroy instance $ID

# 3. Report the host to Vast.ai support
echo "Report host ID to Vast.ai support for investigation"

Incident Type 4: Billing Emergency

# Stop all billing immediately
echo "EMERGENCY: Destroying all instances"
vastai show instances --raw | python3 -c "
import sys, json, subprocess
for inst in json.load(sys.stdin):
    if inst.get('actual_status') in ('running', 'loading'):
        subprocess.run(['vastai', 'destroy', 'instance', str(inst['id'])])
        print(f'Destroyed instance {inst[\"id\"]}')
"

Postmortem Template

## Incident Report
- **Date**: YYYY-MM-DD
- **Duration**: X hours
- **Impact**: N instances affected, $X cost
- **Root cause**: [spot preemption / OOM / disk full / GPU failure]
- **Resolution**: [replaced instance / increased VRAM / expanded disk]
- **Prevention**: [higher reliability filter / checkpoints / auto-recovery]

Output

• Triage script with instant status assessment
• Recovery procedures for 4 incident types
• Emergency billing stop command
• Postmortem template

Error Handling

Incident	MTTR Target	Recovery
Spot preemption	< 10 min	Auto-provision replacement, resume from checkpoint
Training crash	< 5 min	SSH in, diagnose, restart from checkpoint
GPU failure	< 15 min	Destroy instance, provision on different host
Billing emergency	< 1 min	Destroy all instances immediately

Resources

• Vast.ai Status
• Vast.ai CLI

Next Steps

For data handling and security, see vastai-data-handling.

Examples

Auto-recovery script: Run the event poller from vastai-webhooks-events with an auto-recovery handler that provisions a replacement within 5 minutes of preemption.

Kill switch: Keep vastai show instances && vastai destroy instance ALL aliased for emergency billing stops.