Add NIP-11 relay synchronization and group management features
Some checks failed
Go / build (push) Has been cancelled
Go / release (push) Has been cancelled

- Introduced a new `sync` package for managing NIP-11 relay information and relay group configurations.
- Implemented a cache for NIP-11 documents, allowing retrieval of relay public keys and authoritative configurations.
- Enhanced the sync manager to update peer lists based on authoritative configurations from relay group events.
- Updated event handling to incorporate policy checks during event imports, ensuring compliance with relay rules.
- Refactored various components to utilize the new `sha256-simd` package for improved performance.
- Added comprehensive tests to validate the new synchronization and group management functionalities.
- Bumped version to v0.24.1 to reflect these changes.
This commit is contained in:
2025-11-03 18:17:15 +00:00
parent e161d0e4be
commit e56bf76257
83 changed files with 3712 additions and 7417 deletions

691
docs/POLICY_USAGE_GUIDE.md Normal file
View File

@@ -0,0 +1,691 @@
# ORLY Policy System Usage Guide
The ORLY relay implements a comprehensive policy system that provides fine-grained control over event storage and retrieval. This guide explains how to configure and use the policy system to implement custom relay behavior.
## Overview
The policy system allows relay operators to:
- Control which events are stored and retrieved
- Implement custom validation logic
- Set size and age limits for events
- Define access control based on pubkeys
- Use scripts for complex policy rules
- Filter events by content, kind, or other criteria
## Quick Start
### 1. Enable the Policy System
Set the environment variable to enable policy checking:
```bash
export ORLY_POLICY_ENABLED=true
```
### 2. Create a Policy Configuration
Create the policy file at `~/.config/ORLY/policy.json`:
```json
{
"default_policy": "allow",
"global": {
"max_age_of_event": 86400,
"max_age_event_in_future": 300,
"size_limit": 100000
},
"rules": {
"1": {
"description": "Text notes - basic validation",
"max_age_of_event": 3600,
"size_limit": 32000
}
}
}
```
### 3. Restart the Relay
```bash
# Restart your relay to load the policy
sudo systemctl restart orly
```
## Configuration Structure
### Top-Level Configuration
```json
{
"default_policy": "allow|deny",
"kind": {
"whitelist": ["1", "3", "4"],
"blacklist": []
},
"global": { ... },
"rules": { ... }
}
```
### default_policy
Determines the fallback behavior when no specific rules apply:
- `"allow"`: Allow events unless explicitly denied (default)
- `"deny"`: Deny events unless explicitly allowed
### kind Filtering
Controls which event kinds are processed:
```json
"kind": {
"whitelist": ["1", "3", "4", "9735"],
"blacklist": []
}
```
- `whitelist`: Only these kinds are allowed (if present)
- `blacklist`: These kinds are denied (if present)
- Empty arrays allow all kinds
### Global Rules
Rules that apply to **all events** regardless of kind:
```json
"global": {
"description": "Site-wide security rules",
"write_allow": [],
"write_deny": [],
"read_allow": [],
"read_deny": [],
"size_limit": 100000,
"content_limit": 50000,
"max_age_of_event": 86400,
"max_age_event_in_future": 300,
"privileged": false
}
```
### Kind-Specific Rules
Rules that apply to specific event kinds:
```json
"rules": {
"1": {
"description": "Text notes",
"write_allow": [],
"write_deny": [],
"read_allow": [],
"read_deny": [],
"size_limit": 32000,
"content_limit": 10000,
"max_age_of_event": 3600,
"max_age_event_in_future": 60,
"privileged": false
}
}
```
## Policy Fields
### Access Control
#### write_allow / write_deny
Control who can publish events:
```json
{
"write_allow": ["npub1allowed...", "npub1another..."],
"write_deny": ["npub1blocked..."]
}
```
- `write_allow`: Only these pubkeys can write (empty = allow all)
- `write_deny`: These pubkeys cannot write
#### read_allow / read_deny
Control who can read events:
```json
{
"read_allow": ["npub1trusted..."],
"read_deny": ["npub1suspicious..."]
}
```
- `read_allow`: Only these pubkeys can read (empty = allow all)
- `read_deny`: These pubkeys cannot read
### Size Limits
#### size_limit
Maximum total event size in bytes:
```json
{
"size_limit": 32000
}
```
Includes ID, pubkey, sig, tags, content, and metadata.
#### content_limit
Maximum content field size in bytes:
```json
{
"content_limit": 10000
}
```
Only applies to the `content` field.
### Age Validation
#### max_age_of_event
Maximum age of events in seconds (prevents replay attacks):
```json
{
"max_age_of_event": 3600
}
```
Events older than `current_time - max_age_of_event` are rejected.
#### max_age_event_in_future
Maximum time events can be in the future in seconds:
```json
{
"max_age_event_in_future": 300
}
```
Events with `created_at > current_time + max_age_event_in_future` are rejected.
### Advanced Options
#### privileged
Require events to be authored by authenticated users or contain authenticated users in p-tags:
```json
{
"privileged": true
}
```
Useful for private content that should only be accessible to specific users.
#### script
Path to a custom script for complex validation logic:
```json
{
"script": "/path/to/custom-policy.sh"
}
```
See the script section below for details.
## Policy Scripts
For complex validation logic, use custom scripts that receive events via stdin and return decisions via stdout.
### Script Interface
**Input**: JSON event objects, one per line:
```json
{
"id": "event_id",
"pubkey": "author_pubkey",
"kind": 1,
"content": "Hello, world!",
"tags": [["p", "recipient"]],
"created_at": 1640995200,
"sig": "signature"
}
```
Additional fields provided:
- `logged_in_pubkey`: Hex pubkey of authenticated user (if any)
- `ip_address`: Client IP address
**Output**: JSONL responses:
```json
{"id": "event_id", "action": "accept", "msg": ""}
{"id": "event_id", "action": "reject", "msg": "Blocked content"}
{"id": "event_id", "action": "shadowReject", "msg": ""}
```
### Actions
- `accept`: Store/retrieve the event normally
- `reject`: Reject with OK=false and message
- `shadowReject`: Accept with OK=true but don't store (useful for spam filtering)
### Example Scripts
#### Bash Script
```bash
#!/bin/bash
while read -r line; do
if [[ -n "$line" ]]; then
event_id=$(echo "$line" | jq -r '.id')
# Check for spam content
if echo "$line" | jq -r '.content' | grep -qi "spam"; then
echo "{\"id\":\"$event_id\",\"action\":\"reject\",\"msg\":\"Spam detected\"}"
else
echo "{\"id\":\"$event_id\",\"action\":\"accept\",\"msg\":\"\"}"
fi
fi
done
```
#### Python Script
```python
#!/usr/bin/env python3
import json
import sys
def process_event(event):
event_id = event.get('id', '')
content = event.get('content', '')
pubkey = event.get('pubkey', '')
logged_in = event.get('logged_in_pubkey', '')
# Block spam
if 'spam' in content.lower():
return {
'id': event_id,
'action': 'reject',
'msg': 'Content contains spam'
}
# Require authentication for certain content
if 'private' in content.lower() and not logged_in:
return {
'id': event_id,
'action': 'reject',
'msg': 'Authentication required'
}
return {
'id': event_id,
'action': 'accept',
'msg': ''
}
for line in sys.stdin:
if line.strip():
try:
event = json.loads(line)
response = process_event(event)
print(json.dumps(response))
sys.stdout.flush()
except json.JSONDecodeError:
continue
```
### Script Configuration
Place scripts in a secure location and reference them in policy:
```json
{
"rules": {
"1": {
"script": "/etc/orly/policy/text-note-policy.py",
"description": "Custom validation for text notes"
}
}
}
```
Ensure scripts are executable and have appropriate permissions.
## Policy Evaluation Order
Events are evaluated in this order:
1. **Global Rules** - Applied first to all events
2. **Kind Filtering** - Whitelist/blacklist check
3. **Kind-specific Rules** - Rules for the event's kind
4. **Script Rules** - Custom script logic (if configured)
5. **Default Policy** - Fallback behavior
The first rule that makes a decision (allow/deny) stops evaluation.
## Event Processing Integration
### Write Operations (EVENT)
When `ORLY_POLICY_ENABLED=true`, each incoming EVENT is checked:
```go
// Pseudo-code for policy integration
func handleEvent(event *Event, client *Client) {
decision := policy.CheckPolicy("write", event, client.Pubkey, client.IP)
if decision.Action == "reject" {
client.SendOK(event.ID, false, decision.Message)
return
}
if decision.Action == "shadowReject" {
client.SendOK(event.ID, true, "")
return
}
// Store event
storeEvent(event)
client.SendOK(event.ID, true, "")
}
```
### Read Operations (REQ)
Events returned in REQ responses are filtered:
```go
func handleReq(filter *Filter, client *Client) {
events := queryEvents(filter)
filteredEvents := []Event{}
for _, event := range events {
decision := policy.CheckPolicy("read", &event, client.Pubkey, client.IP)
if decision.Action != "reject" {
filteredEvents = append(filteredEvents, event)
}
}
sendEvents(client, filteredEvents)
}
```
## Common Use Cases
### Basic Spam Filtering
```json
{
"global": {
"max_age_of_event": 86400,
"size_limit": 100000
},
"rules": {
"1": {
"script": "/etc/orly/scripts/spam-filter.sh",
"max_age_of_event": 3600,
"size_limit": 32000
}
}
}
```
### Private Relay
```json
{
"default_policy": "deny",
"global": {
"write_allow": ["npub1trusted1...", "npub1trusted2..."],
"read_allow": ["npub1trusted1...", "npub1trusted2..."]
}
}
```
### Content Moderation
```json
{
"rules": {
"1": {
"script": "/etc/orly/scripts/content-moderation.py",
"description": "AI-powered content moderation"
}
}
}
```
### Rate Limiting
```json
{
"global": {
"script": "/etc/orly/scripts/rate-limiter.sh"
}
}
```
### Follows-Based Access
Combined with ACL system:
```bash
export ORLY_ACL_MODE=follows
export ORLY_ADMINS=npub1admin1...,npub1admin2...
export ORLY_POLICY_ENABLED=true
```
## Monitoring and Debugging
### Log Messages
Policy decisions are logged:
```
policy allowed event <id>
policy rejected event <id>: reason
policy filtered out event <id> for read access
```
### Script Health
Script failures are logged:
```
policy rule for kind <N> is inactive (script not running), falling back to default policy (allow)
policy rule for kind <N> failed (script processing error: timeout), falling back to default policy (allow)
```
### Testing Policies
Use the policy test tools:
```bash
# Test policy with sample events
./scripts/run-policy-test.sh
# Test policy filter integration
./scripts/run-policy-filter-test.sh
```
### Debugging Scripts
Test scripts independently:
```bash
# Test script with sample event
echo '{"id":"test","kind":1,"content":"test message"}' | ./policy-script.sh
# Expected output:
# {"id":"test","action":"accept","msg":""}
```
## Performance Considerations
### Script Performance
- Scripts run synchronously and can block event processing
- Keep script logic efficient (< 100ms per event)
- Consider using `shadowReject` for non-blocking filtering
- Scripts should handle malformed input gracefully
### Memory Usage
- Policy configuration is loaded once at startup
- Scripts are kept running for performance
- Large configurations may impact startup time
### Scaling
- For high-throughput relays, prefer built-in policy rules over scripts
- Use script timeouts to prevent hanging
- Monitor script performance and resource usage
## Security Considerations
### Script Security
- Scripts run with relay process privileges
- Validate all inputs in scripts
- Use secure file permissions for policy files
- Regularly audit custom scripts
### Access Control
- Test policy rules thoroughly before production use
- Use `privileged: true` for sensitive content
- Combine with authentication requirements
- Log policy violations for monitoring
### Data Validation
- Age validation prevents replay attacks
- Size limits prevent DoS attacks
- Content validation prevents malicious payloads
## Troubleshooting
### Policy Not Loading
Check file permissions and path:
```bash
ls -la ~/.config/ORLY/policy.json
cat ~/.config/ORLY/policy.json
```
### Scripts Not Working
Verify script is executable and working:
```bash
ls -la /path/to/script.sh
./path/to/script.sh < /dev/null
```
### Unexpected Behavior
Enable debug logging:
```bash
export ORLY_LOG_LEVEL=debug
```
Check logs for policy decisions and errors.
### Common Issues
1. **Script timeouts**: Increase script timeouts or optimize script performance
2. **Memory issues**: Reduce script memory usage or use built-in rules
3. **Permission errors**: Fix file permissions on policy files and scripts
4. **Configuration errors**: Validate JSON syntax and field names
## Advanced Configuration
### Multiple Policies
Use different policies for different relay instances:
```bash
# Production relay
export ORLY_APP_NAME=production
# Policy at ~/.config/production/policy.json
# Staging relay
export ORLY_APP_NAME=staging
# Policy at ~/.config/staging/policy.json
```
### Dynamic Policies
Policies can be updated without restart by modifying the JSON file. Changes take effect immediately for new events.
### Integration with External Systems
Scripts can integrate with external services:
```python
import requests
def check_external_service(content):
response = requests.post('http://moderation-service:8080/check',
json={'content': content}, timeout=5)
return response.json().get('approved', False)
```
## Examples Repository
See the `docs/` directory for complete examples:
- `example-policy.json`: Complete policy configuration
- `example-policy.sh`: Sample policy script
- Various test scripts in `scripts/`
## Support
For issues with policy configuration:
1. Check the logs for error messages
2. Validate your JSON configuration
3. Test scripts independently
4. Review the examples in `docs/`
5. Check file permissions and paths
## Migration from Other Systems
### From Simple Filtering
Replace simple filters with policy rules:
```json
// Before: Simple size limit
// After: Policy-based size limit
{
"global": {
"size_limit": 50000
}
}
```
### From Custom Code
Migrate custom validation logic to policy scripts:
```json
{
"rules": {
"1": {
"script": "/etc/orly/scripts/custom-validation.py"
}
}
}
```
The policy system provides a flexible, maintainable way to implement complex relay behavior while maintaining performance and security.

617
docs/RELAY_TESTING_GUIDE.md Normal file
View File

@@ -0,0 +1,617 @@
# Relay Testing Guide
This guide explains how to use ORLY's comprehensive testing infrastructure for protocol validation, especially when developing features that require multiple relays to test the Nostr protocol correctly.
## Overview
ORLY provides multiple testing tools and scripts designed for different testing scenarios:
- **relay-tester**: Protocol compliance testing against NIP specifications
- **Benchmark suite**: Performance testing across multiple relay implementations
- **Policy testing**: Custom policy validation
- **Integration scripts**: Multi-relay testing scenarios
## Testing Tools Overview
### relay-tester
The primary tool for testing Nostr protocol compliance:
```bash
# Basic usage
relay-tester -url ws://127.0.0.1:3334
# Test with different configurations
relay-tester -url wss://relay.example.com -v -json
```
**Key Features:**
- Tests all major NIP-01, NIP-09, NIP-42 features
- Validates event publishing, querying, and subscription handling
- Checks JSON compliance and signature validation
- Provides both human-readable and JSON output
### Benchmark Suite
Performance testing across multiple relay implementations:
```bash
# Setup external relays
cd cmd/benchmark
./setup-external-relays.sh
# Run benchmark suite
docker-compose up --build
```
**Key Features:**
- Compares ORLY against other relay implementations
- Tests throughput, latency, and reliability
- Provides detailed performance metrics
- Generates comparison reports
### Policy Testing
Custom policy validation tools:
```bash
# Test policy with sample events
./scripts/run-policy-test.sh
# Test policy filter integration
./scripts/run-policy-filter-test.sh
```
## Multi-Relay Testing Scenarios
### Why Multiple Relays?
Many Nostr protocol features require testing with multiple relays:
- **Event replication** between relays
- **Cross-relay subscriptions** and queries
- **Relay discovery** and connection management
- **Protocol interoperability** between different implementations
- **Distributed features** like directory consensus
### Testing Infrastructure
ORLY provides several ways to run multiple relays for testing:
#### 1. Local Multi-Relay Setup
Run multiple instances on different ports:
```bash
# Terminal 1: Relay 1 on port 3334
ORLY_PORT=3334 ./orly &
# Terminal 2: Relay 2 on port 3335
ORLY_PORT=3335 ./orly &
# Terminal 3: Relay 3 on port 3336
ORLY_PORT=3336 ./orly &
```
#### 2. Docker-based Multi-Relay
Use Docker for isolated relay instances:
```bash
# Run multiple relays with Docker
docker run -d -p 3334:3334 -e ORLY_PORT=3334 orly:latest
docker run -d -p 3335:3334 -e ORLY_PORT=3334 orly:latest
docker run -d -p 3336:3334 -e ORLY_PORT=3334 orly:latest
```
#### 3. Benchmark Suite Multi-Relay
The benchmark suite automatically sets up multiple relays:
```bash
cd cmd/benchmark
./setup-external-relays.sh
docker-compose up next-orly khatru-sqlite strfry
```
## Developing Features Requiring Multiple Relays
### 1. Event Replication Testing
Test how events propagate between relays:
```go
// Example test for event replication
func TestEventReplication(t *testing.T) {
// Start two relays
relay1 := startTestRelay(t, 3334)
defer relay1.Stop()
relay2 := startTestRelay(t, 3335)
defer relay2.Stop()
// Connect clients to both relays
client1 := connectToRelay(t, "ws://127.0.0.1:3334")
client2 := connectToRelay(t, "ws://127.0.0.1:3335")
// Publish event to relay1
event := createTestEvent(t)
ok := client1.Publish(event)
assert.True(t, ok)
// Wait for replication/propagation
time.Sleep(100 * time.Millisecond)
// Query relay2 for the event
events := client2.Query(filterForEvent(event.ID))
assert.Len(t, events, 1)
assert.Equal(t, event.ID, events[0].ID)
}
```
### 2. Cross-Relay Subscriptions
Test subscriptions that span multiple relays:
```go
func TestCrossRelaySubscriptions(t *testing.T) {
// Setup multiple relays
relays := setupMultipleRelays(t, 3)
defer stopRelays(t, relays)
clients := connectToRelays(t, relays)
// Subscribe to same filter on all relays
filter := Filter{Kinds: []int{1}, Limit: 10}
for _, client := range clients {
client.Subscribe(filter)
}
// Publish events to different relays
for i, client := range clients {
event := createTestEvent(t)
event.Content = fmt.Sprintf("Event from relay %d", i)
client.Publish(event)
}
// Verify events appear on all relays (if replication is enabled)
time.Sleep(200 * time.Millisecond)
for _, client := range clients {
events := client.GetReceivedEvents()
assert.GreaterOrEqual(t, len(events), 3) // At least the events from all relays
}
}
```
### 3. Relay Discovery Testing
Test relay list events and dynamic relay discovery:
```go
func TestRelayDiscovery(t *testing.T) {
relay1 := startTestRelay(t, 3334)
relay2 := startTestRelay(t, 3335)
defer relay1.Stop()
defer relay2.Stop()
client := connectToRelay(t, "ws://127.0.0.1:3334")
// Publish relay list event (kind 10002)
relayList := createRelayListEvent(t, []string{
"wss://relay1.example.com",
"wss://relay2.example.com",
})
client.Publish(relayList)
// Test that relay discovery works
discovered := client.QueryRelays()
assert.Contains(t, discovered, "wss://relay1.example.com")
assert.Contains(t, discovered, "wss://relay2.example.com")
}
```
## Testing Scripts and Automation
### Automated Multi-Relay Testing
Use the provided scripts for automated testing:
#### 1. relaytester-test.sh
Tests relay with protocol compliance:
```bash
# Test single relay
./scripts/relaytester-test.sh
# Test with policy enabled
ORLY_POLICY_ENABLED=true ./scripts/relaytester-test.sh
# Test with ACL enabled
ORLY_ACL_MODE=follows ./scripts/relaytester-test.sh
```
#### 2. test.sh (Full Test Suite)
Runs all tests including multi-component scenarios:
```bash
# Run complete test suite
./scripts/test.sh
# Run specific package tests
go test ./pkg/sync/... # Test synchronization features
go test ./pkg/protocol/... # Test protocol implementations
```
#### 3. runtests.sh (Performance Tests)
```bash
# Run performance benchmarks
./scripts/runtests.sh
```
### Custom Testing Scripts
Create custom scripts for specific multi-relay scenarios:
```bash
#!/bin/bash
# test-multi-relay-replication.sh
# Start multiple relays
echo "Starting relays..."
ORLY_PORT=3334 ./orly &
RELAY1_PID=$!
ORLY_PORT=3335 ./orly &
RELAY2_PID=$!
ORLY_PORT=3336 ./orly &
RELAY3_PID=$!
# Wait for startup
sleep 2
# Run replication tests
echo "Running replication tests..."
go test -v ./pkg/sync -run TestReplication
# Run protocol tests
echo "Running protocol tests..."
relay-tester -url ws://127.0.0.1:3334 -json > relay1-results.json
relay-tester -url ws://127.0.0.1:3335 -json > relay2-results.json
relay-tester -url ws://127.0.0.1:3336 -json > relay3-results.json
# Cleanup
kill $RELAY1_PID $RELAY2_PID $RELAY3_PID
echo "Tests completed"
```
## Testing Distributed Features
### Directory Consensus Testing
Test NIP-XX directory consensus protocol:
```go
func TestDirectoryConsensus(t *testing.T) {
// Setup multiple relays with directory support
relays := setupDirectoryRelays(t, 5)
defer stopRelays(t, relays)
clients := connectToRelays(t, relays)
// Create trust acts between relays
for i, client := range clients {
trustAct := createTrustAct(t, client.Pubkey, relays[(i+1)%len(relays)].Pubkey, 80)
client.Publish(trustAct)
}
// Wait for consensus
time.Sleep(1 * time.Second)
// Verify trust relationships
for _, client := range clients {
trustGraph := client.QueryTrustGraph()
// Verify expected trust relationships exist
assert.True(t, len(trustGraph.GetAllTrustActs()) > 0)
}
}
```
### Sync Protocol Testing
Test event synchronization between relays:
```go
func TestRelaySynchronization(t *testing.T) {
relay1 := startTestRelay(t, 3334)
relay2 := startTestRelay(t, 3335)
defer relay1.Stop()
defer relay2.Stop()
// Enable sync between relays
configureSync(t, relay1, relay2)
client1 := connectToRelay(t, "ws://127.0.0.1:3334")
client2 := connectToRelay(t, "ws://127.0.0.1:3335")
// Publish events to relay1
events := createTestEvents(t, 100)
for _, event := range events {
client1.Publish(event)
}
// Wait for sync
waitForSync(t, relay1, relay2)
// Verify events on relay2
syncedEvents := client2.Query(Filter{Kinds: []int{1}, Limit: 200})
assert.Len(t, syncedEvents, 100)
}
```
## Performance Testing with Multiple Relays
### Load Testing
Test performance under load with multiple relays:
```bash
# Start multiple relays
for port in 3334 3335 3336; do
ORLY_PORT=$port ./orly &
echo $! >> relay_pids.txt
done
# Run load tests against each relay
for port in 3334 3335 3336; do
echo "Testing relay on port $port"
relay-tester -url ws://127.0.0.1:$port -json > results_$port.json &
done
wait
# Analyze results
# Combine and compare performance across relays
```
### Benchmarking Comparisons
Use the benchmark suite for comparative testing:
```bash
cd cmd/benchmark
# Setup all relay types
./setup-external-relays.sh
# Run benchmarks comparing multiple implementations
docker-compose up --build
# Results in reports/run_YYYYMMDD_HHMMSS/
cat reports/run_*/aggregate_report.txt
```
## Debugging Multi-Relay Issues
### Logging
Enable detailed logging for multi-relay debugging:
```bash
# Enable debug logging
export ORLY_LOG_LEVEL=debug
export ORLY_LOG_TO_STDOUT=true
# Start relays with logging
ORLY_PORT=3334 ./orly 2>&1 | tee relay1.log &
ORLY_PORT=3335 ./orly 2>&1 | tee relay2.log &
```
### Connection Monitoring
Monitor WebSocket connections between relays:
```bash
# Monitor network connections
netstat -tlnp | grep :3334
ss -tlnp | grep :3334
# Monitor relay logs
tail -f relay1.log | grep -E "(connect|disconnect|sync)"
```
### Event Tracing
Trace events across multiple relays:
```go
func traceEventPropagation(t *testing.T, eventID string, relays []*TestRelay) {
for _, relay := range relays {
client := connectToRelay(t, relay.URL)
events := client.Query(Filter{IDs: []string{eventID}})
if len(events) > 0 {
t.Logf("Event %s found on relay %s", eventID, relay.URL)
} else {
t.Logf("Event %s NOT found on relay %s", eventID, relay.URL)
}
}
}
```
## CI/CD Integration
### GitHub Actions Example
```yaml
# .github/workflows/multi-relay-tests.yml
name: Multi-Relay Tests
on: [push, pull_request]
jobs:
test:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: Setup Go
uses: actions/setup-go@v4
with:
go-version: '1.21'
- name: Install dependencies
run: |
sudo apt-get update
sudo apt-get install -y docker.io docker-compose
- name: Run single relay tests
run: ./scripts/relaytester-test.sh
- name: Run multi-relay integration tests
run: |
# Start multiple relays
ORLY_PORT=3334 ./orly &
ORLY_PORT=3335 ./orly &
ORLY_PORT=3336 ./orly &
sleep 3
# Run integration tests
go test -v ./pkg/sync -run TestMultiRelay
- name: Run benchmark suite
run: |
cd cmd/benchmark
./setup-external-relays.sh
docker-compose up --build --abort-on-container-exit
- name: Upload test results
uses: actions/upload-artifact@v3
with:
name: test-results
path: |
cmd/benchmark/reports/
*-results.json
```
## Best Practices
### 1. Test Isolation
- Use separate databases for each test relay
- Clean up resources after tests
- Use unique ports to avoid conflicts
### 2. Timing Considerations
- Allow time for event propagation between relays
- Use exponential backoff for retry logic
- Account for network latency in assertions
### 3. Resource Management
- Limit concurrent relays in CI/CD
- Clean up Docker containers and processes
- Monitor resource usage during tests
### 4. Error Handling
- Test both success and failure scenarios
- Verify error propagation across relays
- Test network failure scenarios
### 5. Performance Monitoring
- Measure latency between relays
- Track memory and CPU usage
- Monitor WebSocket connection stability
## Troubleshooting Common Issues
### Connection Failures
```bash
# Check if relays are listening
netstat -tlnp | grep :3334
# Test WebSocket connection manually
websocat ws://127.0.0.1:3334
```
### Event Propagation Delays
```bash
# Increase wait times in tests
time.Sleep(500 * time.Millisecond)
// Or use polling
func waitForEvent(t *testing.T, client *Client, eventID string) {
timeout := time.After(5 * time.Second)
ticker := time.NewTicker(100 * time.Millisecond)
defer ticker.Stop()
for {
select {
case <-timeout:
t.Fatalf("Event %s not found within timeout", eventID)
case <-ticker.C:
events := client.Query(Filter{IDs: []string{eventID}})
if len(events) > 0 {
return
}
}
}
}
```
### Race Conditions
```go
// Use proper synchronization
var mu sync.Mutex
eventCount := 0
// In test goroutines
mu.Lock()
eventCount++
mu.Unlock()
```
### Resource Exhaustion
```bash
# Limit relay instances in tests
const maxRelays = 3
func setupLimitedRelays(t *testing.T, count int) []*TestRelay {
if count > maxRelays {
t.Skipf("Skipping test requiring %d relays (max %d)", count, maxRelays)
}
// Setup relays...
}
```
## Contributing
When adding new features that require multi-relay testing:
1. Add unit tests for single-relay scenarios
2. Add integration tests for multi-relay scenarios
3. Update this guide with new testing patterns
4. Ensure tests work in CI/CD environment
5. Document any new testing tools or scripts
## Related Documentation
- [POLICY_USAGE_GUIDE.md](POLICY_USAGE_GUIDE.md) - Policy system testing
- [README.md](../../README.md) - Main project documentation
- [cmd/benchmark/README.md](../../cmd/benchmark/README.md) - Benchmark suite
- [cmd/relay-tester/README.md](../../cmd/relay-tester/README.md) - Protocol testing
This guide provides the foundation for testing complex Nostr protocol features that require multiple relay coordination. The testing infrastructure is designed to be extensible and support various testing scenarios while maintaining reliability and performance.