Traffic Management
Load balancing, failover paths, proxy routing, and rollout planning for live services.
Case studies
Infrastructure work proven under real production load.
These projects show how CraftSupport improves reliability, recovery, monitoring, and scaling for active services where downtime has a direct impact on users.
Proof points
Production outcomes
- Monthly active users supported
- 50K
- Peak concurrent players supported
- 1,500+
- Repeated project themes
- BackupsFailoverMonitoringSecurity
What These Projects Required
The same patterns appear across demanding infrastructure: traffic must keep flowing, failures must be observable, and recovery must be planned before it is needed.
Operational Visibility
Dashboards and alerts that make abnormal load, resource pressure, and service health easier to act on.
Access Control
Private access paths, least-privilege controls, firewall tightening, and exposure reduction.
Recovery Planning
Off-site backups, retention strategy, restore documentation, and safer recovery workflows.
Gaming Infrastructure / Minecraft server
CrystalChaos
50K MAU
CraftSupport helped CrystalChaos modernize and stabilize a high-traffic game network after reliability issues with previous infrastructure management.
Audience
50K
monthly active users
Focus
HA
failover and rollouts
Coverage
24/7
monitoring visibility
Challenge
The environment needed safer backups, clearer monitoring, better deployment paths, and stronger access controls without disrupting an active player base.
Technology
ResticHAProxyGrafanaNetdataNetbird
What changed
- Built restic-based off-site backups with deduplication to improve recovery safety and reduce storage overhead.
- Added HAProxy load balancing and automatic failover so updates could be rolled out with less downtime risk.
- Implemented Netdata and Grafana dashboards to surface abnormal load patterns and hardware pressure.
- Moved sensitive access behind Zero Trust controls with Netbird for SSH, databases, and internal services.
Gaming Infrastructure / Minecraft server
Crusalis
1,500+ CCU
CraftSupport re-engineered Crusalis infrastructure to support heavy player concurrency with better traffic distribution and operational confidence.
Peak load
1,500+
concurrent players
Impact
Less
downtime during updates
Scope
Full
network rework
Challenge
The network needed to handle peak load across proxy and hub layers while reducing downtime and preparing for future growth.
Technology
Load balancingProxy routingCustom toolingNetwork design
What changed
- Reworked network infrastructure around more reliable player traffic distribution.
- Built custom load-balancing infrastructure for proxy and hub layers.
- Supported more than 1,500 concurrent players during peak periods.
- Improved stability during updates and reduced operational pressure on the server team.
"Their expertise enabled our server to support over 1,500 concurrent players during peak periods. The result was higher stability, less downtime, and the confidence that we can scale further."
Have a system that needs this level of attention?
Share the service, load profile, and current failure points. We will help identify the highest-impact reliability work first.