High Availability & Active-Active Clustering
For mission-critical enterprise deployments, FaxCore Evolution eV6 supports an Active-Active multi-zone high availability (HA) architecture. This topology eliminates single points of failure across web interfaces, REST APIs, document rendering, and telephony transmission pipelines.
Architectural Topology
The following diagram illustrates the active-active multi-zone deployment utilizing Azure Availability Zones, load balancing, clustered SQL persistence, and high-availability shared storage:
Interactive Traffic Flow & Failover Simulation
Tip
In the online documentation portal, use the interactive simulation below to test traffic flows under normal Active-Active operations, or simulate failover when Availability Zone 1 or Zone 2 is taken offline:
Tier-by-Tier Architecture Breakdown
1. Ingress & Load Balancing Tier
- Azure Traffic Manager / Application Gateway: Provides intelligent DNS routing and health probe monitoring across availability zones.
- Health Probes: Evaluates the
/api/healthor/admin/login.aspxendpoints every 5 seconds. If any node or zone fails health checks, traffic is dynamically diverted to the surviving zone within seconds. - Protocols Handled:
- HTTPS (Port 443): Web Client, Admin Panel, and REST/SOAP API requests.
- SIP / H.323 (Port 5060): Inbound and outbound Fax-over-IP signaling.
2. Multi-Zone Compute Tier (Active-Active)
Nodes are distributed across physically isolated datacenter availability zones with independent power, networking, and cooling:
| Zone | Node | Host IP | Roles & Services Executed |
|---|---|---|---|
| Availability Zone 1 | Node 01 | 10.1.1.4 |
IIS Web Client, Admin Portal, REST API, Render Agent, FaxAgent |
| Node 02 | 10.1.1.5 |
IIS Web Client, Admin Portal, REST API, Render Agent, FaxAgent | |
| Availability Zone 2 | Node 01 | 10.2.1.4 |
IIS Web Client, Admin Portal, REST API, Render Agent, FaxAgent |
| Node 02 | 10.2.1.5 |
IIS Web Client, Admin Portal, REST API, Render Agent, FaxAgent |
3. Shared Resilient Backend Tier
Clustered Database (Azure SQL / SQL AlwaysOn)
- All four application nodes connect to a unified clustered SQL database endpoint (
fxc_V6). - In Azure, utilize Azure SQL Database Hyperscale / Business Critical or SQL Server AlwaysOn Availability Groups with synchronous data replication across zones for zero RPO (Recovery Point Objective).
Resilient Document Store (Azure Files Premium / SAN SMB 3.0)
- The document store repository (
\xdoc\) is hosted on an enterprise SMB 3.0 file share with zone-redundant storage (ZRS). - Every active node references the unified UNC path:
\\storageaccount.file.core.windows.net\faxcore-xdoc\ - All incoming faxes, outbound attachments, and rendered TIFF caches are instantly visible to all compute nodes without manual file replication.
Failover Scenarios & Behavior
Scenario A: Normal Operations (Active-Active)
- Traffic is evenly distributed across both Availability Zone 1 and Availability Zone 2.
- Telephony channels on all 4 nodes actively send and receive faxes.
- Peak throughput scales linearly with the total number of provisioned nodes.
Scenario B: Single Node Failure
- If Node 01 in Zone 1 experiences a failure, the local load balancer removes it from the pool.
- Node 02 in Zone 1 absorbs the local traffic while Zone 2 continues operating unaffected.
Scenario C: Complete Availability Zone Outage
- If an entire datacenter zone loses power or network connectivity:
- The Traffic Manager / Gateway detects failed health probes across Zone 1.
- Zone 1 is immediately marked Offline.
- Ingress routing automatically diverts 100% of API, Web, and SIP traffic to Zone 2.
- Zone 2 nodes scale processing capacity and process outbound queues seamlessly.
- When Zone 1 recovers, health checks pass and nodes automatically rejoin the active cluster.