Best Practices for Gluesync

This guide outlines best practices for configuring and using Gluesync effectively, based on real-world experience and customer deployments.

Connection Management

Connection Pooling

  • Default maximum of 100 connections per agent

  • One connection per entity (table/mapping) by default

  • For deployments with:

    • < 100 entities: Default settings work well

    • > 100 entities: Consider connection grouping

    • Large deployments: Adjust max concurrency settings

For detailed guidance on connection pool sizing and optimization, refer to our Connection Pool Management Guide.

Monitor source database load when using many connections. Adjust settings if needed to prevent overload.

Connection Grouping

From the agent advanced settings page you can change the connection grouping properties

  • Use when:

    • Source system has connection limits

    • Need to reduce resource usage

    • Managing many entities

  • Benefits:

    • Reduced connection overhead

    • Better resource utilization

    • Improved stability

Performance Optimization

Polling Intervals

Setting Use Case Impact

15ms

Ultra-low latency (banking, telecom)

Highest resource usage

100ms

Default - good balance

Recommended for most cases

1000ms+

Low priority updates

Minimal resource impact

Start with default 100ms polling and adjust based on: * Business requirements * Source system capacity * Network latency

Memory Management

  • Transaction Message Size:

    • Default: 1MB per message

    • Adjust based on:

      • Available RAM

      • Data volume

      • Update frequency

  • Buffer Management:

    • System auto-adjusts based on target performance

    • Allows accumulation during target slowdowns

    • Prevents synchronization issues

Data Handling

Character Handling

  • Enable automatic trimming for VARCHAR fields when needed:

From the agent advanced settings page you can change the trimming properties

Bidirectional Sync

  • Always enable recursion protection:

From the agent advanced settings page you can enable resursion protection

  • Prevents:

    • Infinite update loops

    • Data corruption

    • Unnecessary replication

Deployment Patterns

Agent Placement

  • Place agents close to data sources/targets

  • Options:

    • Same deployment as Core Hub

    • Near source/target systems

    • Hybrid deployment

Consider network latency and security requirements when placing agents.

Multi-Target Configuration

  • Available deployment modes:

    • Single target (stable)

    • Multi-target with mirroring (beta)

  • Mirroring priorities:

    • Strong consistency: All targets sync together

    • Eventual consistency: Targets sync independently

Monitoring and Maintenance

Regular Checks

  • Monitor notification hub for:

    • System status

    • Error messages

    • Performance issues

  • Review connection status

  • Check resource usage

Error Handling

  • Synchronous errors:

    • Appear immediately

    • Allow retry

    • Usually configuration-related

  • Asynchronous errors:

    • Appear in notification hub

    • May require investigation

    • Often runtime-related

Security Best Practices

TLS Configuration

  • Enable TLS when possible

  • Configure trust stores properly

  • Use secure connections for production

Authentication

  • Use strong credentials

  • Rotate API keys regularly

  • Implement least privilege access

Scaling Guidelines

Vertical Scaling

  • Increase agent resources for:

    • Higher transaction volumes

    • More concurrent operations

    • Larger data sets

Horizontal Scaling

  • Add agents for:

    • New data sources

    • Geographic distribution

    • Load balancing

Always test configuration changes in a non-production environment first.

Common Use Cases

Real-time Replication

  • Use low polling intervals

  • Monitor latency

  • Enable appropriate buffering

Batch Processing

  • Increase transaction message size

  • Adjust polling intervals

  • Configure appropriate timeouts

Event Streaming

  • Configure appropriate TTL

  • Use proper key composition

  • Enable filtering as needed

Support and Troubleshooting

Best Practices for Support

  • Keep logs accessible

  • Document configuration changes

  • Monitor system regularly

Common Issues

  • Connection problems:

    • Check network connectivity

    • Verify credentials

    • Review firewall rules

  • Performance issues:

    • Review polling intervals

    • Check connection pooling

    • Monitor resource usage