.png){kind=logo}
Installation guides for multi-node, high-availability configurations suitable for production environment workloads and large data volumes.
High Availability Failure Tolerances
AgileSec clusters must maintain a quorum to tolerate node failure. A quorum means a majority of AgileSec voting member nodes are operating and communicative.
Voting member nodes include all nodes with MongoDB, Kafka, and OpenSearch services: Backend Nodes, Primary Frontend Node (frontend-1), Coordinator Node. High-availability clusters must always have an even number of Backend Nodes to maintain an odd-number of total voting member nodes for quorum considerations.
|
Failure Tolerance How many nodes can fail without service interruption |
Minimum Configuration |
Guide Recommendation |
|---|---|---|
|
1 Backend Node |
3 Nodes: 2 Backend, 1 Frontend |
|
|
1 Node (any type) |
4 Nodes: 2 Backend, 2 Frontend |
New installation: HA Scenario 1: Four-Node Installation for New Cluster (2 Backends, 2 Frontends) Upgrade from 3-node: HA Scenario 2: Adding a New Frontend to an Existing Cluster |
|
Up to 2 Nodes (any type) Note: If the Frontend Node goes down, data will still be intact but external access will be restricted. |
5 Nodes:
|
HA Scenario 1: Four-Node Installation for New Cluster (2 Backends, 2 Frontends) or HA Scenario 3: Multi-Region Stretch Cluster with Coordinator Node |
|
1 Data Center Outage |
7 Nodes across an odd number of data centers (3+ data centers):
Note: Ideally (non-minimum), each data center will have at least 3 nodes for full failover protection:
|
HA Scenario 3: Multi-Region Stretch Cluster with Coordinator Node |
High-Availability Guides
The following guides are available to configure High Availability installations for both testing and production purposes. All nodes in all scenarios are expected to meet the requirements defined in Multi-Node Linux Install Prerequisites.
Recommendation: Review the basic 3-node Multi-Node Linux Installation guide first.
-
Multi-Node Linux Install Prerequisites Prerequisites and requirements for Multi-Node and High-Availability AgileSec Linux installations, including system requirements, firewall rules, etc.
-
Multi-Node Linux Installation Deploy a basic scalable, high-availability, three-node cluster with 2 Backend Nodes and 1 Frontend Node to support production environment workloads and large data volumes. This is suitable for most production use cases and can be scaled up later.
-
HA Scenario 1: Four-Node Installation for New Cluster (2 Backends, 2 Frontends) Deploy a new, highly available (no single point of failure) multi-node cluster install with two Backend Nodes and two Frontend Nodes to support high-reliability production environment workloads and large data volumes.
-
HA Scenario 2: Adding a New Frontend to an Existing Cluster Add capacity to an existing cluster with an additional Frontend Node to eliminate points of failure and achieve higher traffic availability.
-
HA Scenario 3: Multi-Region Stretch Cluster with Coordinator Node Add a Coordinator Node to Stretch Clusters to maintain majority during an availability zones (AZ) or data centers (DC) failure. This ensures the Stretch Cluster will remain operational if one data-center fails, ensuring true high availability.
-
Additional High Availability Install Configuration Options Optional post-installation configurations to reduce latency and increase high availability.