Instance Management BETA
Monitor and manage consumer instances running on your infrastructure to ensure optimal performance and reliability.
As a provider, you host isolated compute instances for consumers. While SLYD handles provisioning
and access control, understanding instance management helps you optimize resource utilization,
troubleshoot issues, and maximize uptime for better earnings.
Instance Lifecycle
Understanding the instance lifecycle helps you manage resources effectively.
Pending
Instance requested, resources being allocated
→
Starting
Container provisioning and initialization
→
Running
Active and accessible by consumer
→
Stopping
Shutdown initiated, cleaning up
→
Terminated
Resources released and cleaned
Monitoring Active Instances
Use the provider dashboard and CLI tools to monitor instances running on your servers.
Dashboard View
The instances panel shows:
- Instance ID and status
- Consumer information
- Resource allocation (CPU/RAM/Storage)
- Running time and cost
- Network usage statistics
- Health indicators
CLI Monitoring
# List all active instances
slyd-provider instances list
# Get detailed instance info
slyd-provider instances info i-1234567890
# Monitor instance resources
slyd-provider instances stats i-1234567890
# Watch real-time metrics
slyd-provider instances watch
Resource Allocation & Limits
SLYD automatically manages resource allocation, but understanding the process helps optimize your infrastructure.
How Resources Are Allocated
CPU Allocation
- Dedicated cores per instance type
- CPU pinning for performance
- Fair scheduling between instances
- Overcommit ratios configurable
Memory Allocation
- Guaranteed memory per instance
- No overcommit on memory
- Swap disabled by default
- Memory limits enforced
Storage Allocation
- Thin provisioning supported
- Storage quotas enforced
- Automatic cleanup on termination
- Snapshot support available
Network Resources
- Isolated network namespaces
- Bandwidth limits per instance
- DDoS protection included
- IPv4 and IPv6 support
Supported Instance Types
Your servers may host various instance types based on your hardware capabilities.
| Instance Type |
vCPUs |
Memory |
Storage |
Network |
Typical Use Case |
|---|
| Micro |
1 |
1 GB |
10 GB |
100 Mbps |
Development, testing |
| Small |
2 |
4 GB |
40 GB |
250 Mbps |
Web servers, small apps |
| Medium |
4 |
8 GB |
80 GB |
500 Mbps |
Databases, app servers |
| Large |
8 |
16 GB |
160 GB |
1 Gbps |
Data processing, analytics |
| GPU-Enabled |
8 |
32 GB |
200 GB |
10 Gbps |
AI/ML, rendering |
Monitor instance performance to ensure quality service and identify potential issues.
Troubleshooting Instance Issues
Quick guide to diagnose and resolve common instance problems.
Instance Won't Start
Diagnostic Steps:
# Check instance logs
slyd-provider instances logs i-1234567890
# Verify resource availability
slyd-provider resources available
# Check for errors
slyd-provider instances diagnose i-1234567890
Common Causes:
- Insufficient resources available
- Storage pool full
- Network configuration issues
- Container image corruption
Poor Performance
Investigation Commands:
# Check resource contention
slyd-provider instances stats --all
# Identify noisy neighbors
slyd-provider instances top
# Review system load
slyd-provider system stats
Resolution Steps:
- Adjust CPU overcommit ratios
- Enable CPU pinning for instances
- Check for storage I/O bottlenecks
- Review network congestion
Instance Security & Isolation
SLYD ensures strong security and isolation between instances on your servers.
Security Measures
Container Isolation
- Namespace separation (PID, Network, Mount)
- Cgroup resource limits
- Seccomp profiles enabled
- AppArmor/SELinux enforcement
Network Security
- Isolated virtual networks
- Firewall rules per instance
- DDoS protection
- Encrypted tunnel access only
Instance Maintenance Procedures
Handle server maintenance with minimal impact on running instances.
Graceful Maintenance Process
1
Schedule Maintenance
Announce maintenance window to affected consumers
slyd-provider maintenance schedule --start "2024-02-01 02:00" --duration 2h
2
Stop New Instances
Prevent new instance launches
slyd-provider maintenance start --no-new-instances
3
Migrate if Possible
Live migrate instances to other servers (if available)
slyd-provider instances migrate --all --target-server srv-02
4
Complete Maintenance
Perform updates and resume normal operations
slyd-provider maintenance complete
Instance Management Best Practices
Follow these practices to provide the best experience for consumers.
Regular Monitoring: Check instance health metrics at least twice daily
Resource Planning: Maintain 20% resource headroom for peak usage
Quick Response: Address instance issues within 15 minutes when possible
Documentation: Keep notes on recurring issues and solutions
Communication: Notify consumers promptly about any impacts
