These are the metrics that matter. Each one has a healthy range and guidance on what to do when it's not healthy.
Email Flow Metric Healthy Range Warning Critical Action Mail queue depth < 100 > 500 > 5,000 Check Postfix, increase workers Bounce rate < 2% > 3% > 5% Clean lists, check blacklists Spam score (avg outbound) < 3 > 5 > 8 Check content, verify DKIM /SPF Delivery latency (p95) < 5s > 15s > 60s Check DNS, queue congestion Deferred messages < 50 > 200 > 1,000 Check remote server status, rate limits Rejection rate (inbound) < 20% > 40% > 60% Review Rspamd thresholds
API Metric Healthy Range Warning Critical Action Response time (p95) < 500ms > 1s > 5s Check DB queries, add caching Error rate (5xx) < 0.1% > 1% > 5% Check logs, DB connection Requests per second < 80% capacity > 80% > 95% Scale API instances Active connections < 80% of limit > 90% > 95% Increase limits, check for leaks
Database (MySQL) Metric Healthy Range Warning Critical Action Connections used < 70% of max > 80% > 95% Increase max_connections, pool Slow queries/min 0-2 > 10 > 50 Add indexes, optimize queries InnoDB buffer hit rate > 99% < 95% < 90% Increase buffer_pool_size Replication lag (if replica) < 1s > 5s > 30s Check replica load Table lock waits 0 > 5/min > 20/min Optimize queries, batch deletes Disk usage < 70% > 80% > 90% Clean old data, expand disk
Redis Metric Healthy Range Warning Critical Action Memory usage < 70% of max > 80% > 95% Increase maxmemory, tune eviction Hit rate > 90% < 80% < 60% Check eviction policy, increase memory Connected clients < 500 > 800 > 950 Check for connection leaks Evictions/sec 0 > 10 > 100 Increase memory Command latency (p99) < 1ms > 5ms > 50ms Check server load, slow commands
System Resources Metric Healthy Range Warning Critical Action CPU usage < 70% > 80% > 95% Scale up, optimize services Memory usage < 80% > 85% > 95% Increase RAM, tune services Disk usage < 70% > 80% > 90% Clean up, expand storage Disk I/O wait < 5% > 10% > 20% Faster disk (SSD/NVMe), reduce writes Network throughput < 70% of link > 80% > 95% Upgrade bandwidth Load average (per core) < 1.0 > 2.0 > 5.0 Find bottleneck, scale up
SSL Certificates Metric Healthy Range Warning Critical Action Days until expiry > 30 < 14 < 3 Check cert_manager, manual renewal
Workers Metric Healthy Range Warning Critical Action Health check status UP — DOWN Restart, check logs Task processing rate Stable Declining Zero Check dependencies Error rate < 1% > 5% > 20% Check logs, fix root cause Queue backlog < 100 > 500 > 5,000 Increase worker concurrency
Deliverability Metric Healthy Range Warning Critical Action Google domain reputation High Medium Low/Bad Reduce volume, clean lists Blacklist listings 0 1 > 2 Investigate, request delisting SPF pass rate> 99% < 95% < 90% Fix SPF record DKIM pass rate> 99% < 95% < 90% Check DKIM keys DMARC pass rate> 95% < 90% < 80% Fix alignment issues
How to Check These Metrics Quick command-line checks: Bash # Queue depth
exec -it postfix postqueue -p | tail -1
# MySQL connections
exec -it mysql mysql -u root -p" $DB_ROOT_PASSWORD " -e "SHOW STATUS LIKE 'Threads_connected';"
# Redis memory
exec -it redis redis-cli info memory | grep used_memory_human
# Disk usage
-h /
# System load
Via Prometheus (if monitoring is set up): PromQL # Queue depth
mailyte_mail_queue_size
# Bounce rate
rate ( mailyte_emails_bounced_total [ 1h ] ) / rate ( mailyte_emails_sent_total [ 1h ] )
# API latency p95
histogram_quantile ( 0.95 , rate ( mailyte_api_request_duration_seconds_bucket [ 5m ] ))
# MySQL connections used %
mysql_global_status_threads_connected / mysql_global_variables_max_connections