NGINX: The Simplest Load Balancing Cluster with Failover

Virtual IP address in a farm cluster

On the previous figure, the NGINX application is running on the 3 servers (3 is an example, it can be 2 or more). Users are connected to a virtual IP address.
The virtual IP address is configured locally on each server in the farm cluster.
The input traffic to the virtual IP address is received by all the servers and split among them by a network filter inside each server's kernel.
SafeKit detects hardware and software failures, reconfigures network filters in the event of a failure, and offers configurable application checkers and recovery scripts.

Load balancing in a network filter

The network load balancing algorithm inside the network filter is based on the identity of the client packets (client IP address, client TCP port). Depending on the identity of the client packet input, only one filter in a server accepts the packet; the other filters in other servers reject it.
Once a packet is accepted by the filter on a server, only the CPU and memory of this server are used by the NGINX application that responds to the request of the client. The output messages are sent directly from the application server to the client.
If a server fails, the farm heartbeat protocol reconfigures the filters in the network load balancing cluster to re-balance the traffic on the remaining available servers.

Stateful or stateless applications

With a stateful NGINX application, there is session affinity. The same client must be connected to the same server on multiple TCP sessions to retrieve its context on the server. In this case, the SafeKit load balancing rule is configured on the client IP address. Thus, the same client is always connected to the same server on multiple TCP sessions. And different clients are distributed across different servers in the farm.
With a stateless NGINX application, there is no session affinity. The same client can be connected to different servers in the farm on multiple TCP sessions. There is no context stored locally on a server from one session to another. In this case, the SafeKit load balancing rule is configured on the TCP client session identity. This configuration is the one which is the best for distributing sessions between servers, but it requires a TCP service without session affinity.

Comparison of SafeKit with Traditional High Availability (HA) Clusters

How does SafeKit compare to traditional High Availability (HA) cluster solutions?

This comparison highlights the fundamental differences between SafeKit and traditional High Availability (HA) cluster solutions like Failover Clusters, Virtualization HA, and SQL Always-On. SafeKit is designed as a low-complexity, software-only solution for generic application redundancy, contrasting with the high complexity and specific storage requirements (shared storage, SAN) typical of traditional HA mechanisms.

Comparison of SafeKit with traditional High Availability (HA) clusters
Solutions	Complexity	Comments
Failover Cluster (Microsoft)	High	Specific Storage (shared storage, SAN)
Virtualization (VMware HA)	High	Specific Storage (shared storage, SAN, vSAN)
SQL Always-On (Microsoft)	High	Only SQL is redundant, requires SQL Enterprise Edition
Evidian SafeKit	Low	Simplest, generic and software-only. Unsuitable for large data replication.

SafeKit's Advantage in Application Redundancy

SafeKit achieves its low-complexity High Availability through a simple, software-based mirroring mechanism that eliminates the need for expensive, dedicated hardware like a SAN (Storage Area Network). This makes it a highly accessible solution for quickly implementing application redundancy without complex infrastructure changes.

Architectural Differentiators: SafeKit Software-Defined vs. Hardware HA Clusters

Which High Availability Architecture Is Right for You: SafeKit Software Clustering or Traditional Hardware Clustering?

Choosing the right High Availability (HA) solution is critical for ensuring business continuity and minimizing downtime. This comparison provides a direct, technical review of two major architectural approaches: SafeKit's software-defined, shared-nothing clustering versus Traditional HA methods that typically rely on hardware, shared disks (like a SAN), and complex configurations. These distinctions cover deployment simplicity, data replication methods, recovery speed (RTO/RPO), and operational complexity. The tables below detail the core differences across key high availability topics.

Software Clustering vs. Hardware Clustering More info >
SafeKit (Software Clustering)	Hardware Clustering
A simple software cluster with the SafeKit package just installed on two servers	Complex hardware clustering with external storage or network load balancers

Shared Nothing vs. a Shared Disk Cluster More info >
SafeKit (Shared Nothing Cluster)	Shared Disk Cluster
SafeKit is a shared-nothing cluster: easy to deploy even in remote sites	A shared disk cluster is complex to deploy

Application High Availability vs. Full Virtual Machine High Availability More info >
Application High Availability	Virtual Machine High Availability
Application HA supports hardware failure and software failure with application checkers. Quick recovery time by restarting only the application (RTO around 1 mn or less). Application HA requires to define restart scripts per application and folders to replicate (SafeKit application modules).	Full virtual machines HA supports hardware failure and some software failures like a frozen VM. VM reboot on failure and recovery time depending on the OS reboot. No restart scripts to define with full virtual machines HA (SafeKit hyperv.safe or kvm.safe modules). Hypervisors are active/active with just multiple virtual machines.

High Availability vs. Fault Tolerance More info >
SafeKit (High Availability)	Fault Tolerance
No dedicated server with SafeKit. Each server can be the failover server of the other one. Software failure with restart in another OS environment. Smooth upgrade of application and OS possible server by server (version N and N+1 can coexist)	Secondary server dedicated to the execution of the same application synchronized at the instruction level. Software exception on both servers at the same time. Smooth upgrade not possible Specific fault-tolerant hardware or hypervisors

Synchronous Replication vs. Asynchronous Replication More info >
SafeKit (Synchronous Replication)	Asynchronous Replication
SafeKit implements real-time synchronous replication with no data loss in case of failure Prerequisite for high availability	With asynchronous replication, there is data loss on failure Not suited for high availability but for backup solutions

Byte-level File Replication vs. Block-level Disk Replication More info >
SafeKit (Byte-level File Replication)	Block-level Disk Replication
SafeKit implements real-time byte-level file replication and is simply configured with application directories to replicate even in the system disk	Block-level disk replication is complex to configure and requires to put application data in a special disk

Heartbeat, Failover, and Quorum to Avoid 2 Master Nodes More info >
SafeKit	Traditional HA
To avoid 2 masters, SafeKit proposes a simple split brain checker configured on a router	To avoid 2 masters, other clusters require a complex configuration with a third machine, a special quorum disk, a special interconnect

Virtual IP Address: Primary/Secondary, Network Load Balancing, Failover More info >
SafeKit	Traditional HA
No dedicated proxy servers and no special network configuration are required in a SafeKit cluster for virtual IP addresses	Special network configuration is required in other clusters for virtual IP addresses. Note that SafeKit offers a health check adapted to load balancers

Summary and Key Takeaways for High Availability

The architectural choice between software clustering (like SafeKit) and hardware clustering (traditional shared-disk/SAN) significantly impacts deployment complexity, operational costs, and recovery effectiveness. The key takeaway from this comparison is the shift toward shared-nothing, application-level HA which prioritizes rapid application recovery (low RTO) and deployment flexibility (even across remote sites), often resulting in a more streamlined and resilient solution than highly complex, hardware-dependent cluster configurations. For maximum business continuity with simplified management, evaluating a software-based approach is essential.

Key Differentiators of the SafeKit Mirror Cluster

What are the key features and advantages of the SafeKit Mirror Cluster for High Availability (HA)?

Traditional clustering solutions often suffer from high complexity, reliance on costly shared storage (SAN), and difficult management. The SafeKit Mirror Cluster is engineered to solve these problems by offering a software-defined, shared-nothing architecture. This approach not only ensures zero data loss (RPO=0) through synchronous file replication but also dramatically reduces deployment time and expense. The following table breaks down the core features that make SafeKit a flexible, cost-effective, and unified solution for mission-critical applications on Windows and Linux.

Evidian SafeKit Mirror Cluster Core Features and Benefits
Feature Category & Advantage	Detailed Benefit and Mechanism
3 products in 1 More info >	The SafeKit high availability software saves costs on Windows and Linux by eliminating the need for: external shared or replicated storage, load balancing boxes, enterprise editions of OS and databases. SafeKit includes all clustering features: synchronous real-time file replication, monitoring of server / network / software failures, automatic application restart, and virtual IP address switching to reroute clients upon failure.
Very simple configuration More info >	The cluster configuration is very simple and achieved by means of application modules. New services and replicated directories can be easily added to an existing module. All cluster configuration is managed using a simple centralized web administration console. There is no domain controller or Active Directory configuration required, unlike with Microsoft cluster solutions.
Synchronous replication More info >	The real-time replication is synchronous, guaranteeing no data loss upon failure. This is not the case with asynchronous replication, which carries a risk of data loss.
Fully automated failback More info >	After a server reboots following a failure, the replication failback procedure is fully automatic, and the failed server reintegrates the cluster without stopping the application on the remaining server. This is not the case with most replication solutions (especially database-level replication), where manual operations are required, and the application may even need to be stopped during resynchronization.
Replication of any type of data More info >	Replication works for databases as well as for any other files that need to be replicated. This is not the case for replication performed only at the database level.
File replication vs disk replication More info >	Replication is based on file directories that can be located anywhere (even on the system disk). This is not the case with disk replication, which requires special application configuration to place data on a dedicated disk.
File replication vs shared disk More info >	The servers can be placed in two remote sites (Shared-Nothing architecture). This is not the case with shared disk solutions, which require close proximity.
Remote sites and virtual IP address More info >	All SafeKit clustering features work for two servers in remote sites. Replication requires an extended LAN type network (latency affects synchronous replication performance; bandwidth affects resynchronization after failure). If both servers are connected to the same IP network through an extended LAN, the SafeKit virtual IP address works with Layer 2 rerouting. If both servers are connected to two different IP networks, the virtual IP address can be configured at the load balancer level using SafeKit's "health check."
Quorum and split brain More info >	The solution works with only two servers. For the quorum (network isolation between sites), a simple split brain checker to a router is offered to support a single execution of the critical application. This is not the case for most clustering solutions where a third server or witness is required for the quorum.
Active/active cluster More info >	The secondary server is not dedicated solely to the restart of the primary server. The cluster can be active-active by running two different mirror modules concurrently. This is not the case with a fault-tolerant system where the secondary is dedicated to running the exact same application synchronized at the instruction level.
Uniform high availability solution More info >	SafeKit implements a mirror cluster (replication and failover) and a farm cluster (load balancing and failover). Thus, a N-tiers architecture can be made highly available and load balanced with the same solution on Windows and Linux (same installation, configuration, administration via the SafeKit console or command line interface). This unified approach is unique. This is not the case with architectures mixing different technologies for load balancing, replication, and failover.
RTO / RPO More info >	SafeKit implements quick application restart in case of failure: around 1 minute or less (low RTO). Quick application restart is not ensured with full virtual machine replication. In case of hypervisor failure, a full VM must be rebooted on a new hypervisor, with a recovery time dependent on the OS reboot (as with VMware HA or Hyper-V cluster).

Summary of SafeKit Mirror Cluster Benefits for High Availability

In summary, the SafeKit mirror cluster delivers a compelling high availability solution through its shared-nothing architecture and synchronous file replication. By offering a unified platform that bundles replication, monitoring, and failover/failback mechanisms, it successfully addresses critical enterprise needs like zero data loss (RPO=0) and fast Recovery Time Objectives (RTO) of around 1 minute or less. Its simplicity, lack of dependency on expensive SANs or enterprise OS editions, and ability to handle remote sites and active-active configurations make it a highly cost-effective and flexible alternative to complex traditional cluster solutions.

Key Differentiators of the SafeKit Farm Cluster

What are the key differentiators of the SafeKit Farm Cluster for load balancing and failover?

The SafeKit Farm Cluster is a high availability solution specifically designed for scalable application environments where load distribution and rapid failover are essential. Unlike traditional methods that require dedicated hardware load balancers or complex network configurations, SafeKit provides an integrated, software-defined clustering solution installed directly on the application servers. The following table details the core features and unique advantages of the SafeKit Farm Cluster, focusing on how it simplifies network load balancing and ensures continuous service availability across Windows and Linux platforms.

Key Differentiators of SafeKit Farm Cluster with Load Balancing and Failover
Feature Category & Advantage	Detailed Benefit and Mechanism
No load balancer or dedicated proxy servers or special multicast Ethernet address More info >	The solution does not require load balancers or dedicated proxy servers above the farm for implementing load balancing. SafeKit is installed directly on the application servers in the farm. The load balancing is based on a standard virtual IP address/Ethernet MAC address and is working with physical servers or virtual machines on Windows and Linux without special network configuration This is not the case with network load balancers This is not the case with dedicated proxies on Linux This is not the case with a specific multicast Ethernet address on Windows
All clustering features More info >	The solution includes all clustering features: virtual IP address, load balancing on client IP address or on sessions, monitoring of server / network / software failures, automatic application restart with a quick revovery time and a replication option with a mirror module This is not the case with other load balancing solutions. They are able to make load balancing but they do not include a full clustering solution with restart scripts and automatic application restart in case of failure. They do not offer a replication option The cluster configuration is very simple and made by means of application modules. There is no domain controller or active directory to configure on Windows. The solution works on Windows and Linux
Remote sites and virtual IP address More info >	If servers are connected to the same IP network through an extended LAN between remote sites, the virtual IP address of SafeKit is working with load balancing at level 2 If servers are connected to different IP networks between remote sites, the virtual IP address can be configured at the level of a load balancer with the help of the SafeKit health check. Thus you can implement load balancing but also all the clustering features of SafeKit, in particular monitoring and automatic recovery of the critical application on application servers
Uniform high availability solution More info >	SafeKit implements a farm cluster with load balancing and failover. But it implements also a mirror cluster with replication and failover. Thus a N-tiers architecture can be made highly available and load balanced with the same solution on Windows and Linux (same installation, configuration, administration with the SafeKit console or with the command line interface). This is unique on the market This is not the case with an architecture mixing different technologies for load balancing, replication and failover

Summary of SafeKit Farm Cluster Benefits for Load Balancing

In conclusion, the SafeKit Farm Cluster provides a unified, software-based approach to load balancing and high availability that dramatically lowers complexity and cost. By embedding load balancing and failover directly into the application server layer using a standard virtual IP address, it avoids the need for external network hardware (load balancers or proxies) and specialized multicast configurations. This integrated approach, coupled with its ability to combine with the mirror cluster for full N-tiers HA, makes SafeKit a uniquely simple and comprehensive solution for achieving scalable and resilient application delivery across diverse environments.

SafeKit HA Comparison: Virtual Machine Level vs. Application Level

What are the fundamental differences between SafeKit's VM-based and Application-based High Availability?

Understanding the architecture of High Availability (HA) is crucial for selecting the right redundancy solution. SafeKit offers two primary approaches to ensure business continuity: Virtual Machine HA (VM HA) and Application HA. While both methods provide automatic failover capabilities, they differ significantly in their scope, data replication mechanisms, recovery speed, and platform compatibility. This comparison breaks down these differences to help identify the optimal strategy for specific IT environments, whether the focus is on broad virtualization support or granular, high-speed application recovery.

Feature Comparison: SafeKit VM HA vs. SafeKit Application HA Clustering
Comparison Feature	VM HA with SafeKit Hyper-V or KVM Module	Application HA with SafeKit Application Modules
Deployment Diagram
Failover Scope	SafeKit inside 2 hypervisors: replication and failover of the full VM.	SafeKit inside 2 virtual or physical machines: replication and failover at the application level.
Data Replicated	Replicates more data (Application + Operating System).	Replicates only application data, leading to smaller data volumes.
Recovery Process & Speed (RTO)	Reboot of VM on hypervisor 2 if hypervisor 1 crashes. Recovery time depends on the OS reboot. VM checker and failover mechanism.	Quick recovery time with restart of App on OS2 if server 1 crashes. Typically around 1 minute or less (low RTO). Application checker and software failover.
Configuration	Generic solution for any application / OS running in the VM. It does not require a technical understanding of the application installed within the VM. It is the best solution if you do not know how the application works. You only need to define the location of the VM files.	It requires a technical understanding of the application itself. Which services need to be restarted. The specific application folders that need real-time replication. The configuration of a virtual IP address for failover.
Platform Compatibility	Works with Windows/Hyper-V and Linux/KVM but is not compatible with VMware.	Platform agnostic; works with physical or virtual machines, cloud infrastructure, and any hypervisor, including VMware.

Final Recommendation: VM HA for Generality vs. Application HA for Low RTO

In summary, choosing between SafeKit's VM HA and Application HA depends on the priority. VM HA is a generic solution ideal for environments standardized on Hyper-V or KVM, offering redundancy for the entire operating system stack, though with a potentially longer Recovery Time Objective (RTO) due to the VM reboot. Conversely, Application HA provides superior agility and platform agnosticism, including support for VMware, resulting in a much lower RTO by focusing solely on critical application data replication and restart. For the lowest RTO and maximum deployment flexibility, Application HA is the optimal SafeKit choice.

VM High Availability: SafeKit's SAN-Less vs. Hyper-V/VMware HA

What is the difference between SafeKit VM High Availability and Traditional Shared Storage Clusters (Hyper-V Cluster and VMware HA)?

Choosing the right Virtual Machine (VM) High Availability (HA) solution is critical for maintaining business continuity. IT organizations often face a primary architectural choice: utilizing traditional HA clusters, which require expensive Shared Storage like a Storage Area Network (SAN), or adopting modern, software-based solutions that eliminate this dependency. This comparison provides a detailed breakdown of the key differences between the diskless, real-time replication approach of SafeKit (using its Hyper-V and KVM modules) and the classical shared-disk clustering architectures typified by Microsoft Hyper-V Cluster and VMware HA.

Architectural Comparison: SafeKit Non-Shared Storage HA vs. Traditional SAN-Based Clusters
SafeKit (with Hyper-V or KVM Module)	Microsoft Hyper-V Cluster & VMware HA (Traditional)

No shared disk required - uses synchronous real-time replication instead, ensuring no data loss.	Requires shared disk and a specific external disk bay (SAN).
Supports Remote Sites without requiring SAN replication across locations.	Remote sites typically require replicating disk bays across a complex SAN setup.
No specific IT skill is required to configure the system (using hyperv.safe and kvm.safe).	Requires specific, high-level IT skills to configure the cluster and SAN infrastructure.
Note that the Hyper-V/SafeKit and KVM/SafeKit solutions are limited to replication and failover of 32 VMs.	Note that the Hyper-V built-in replication (Hyper-V Replica) does not qualify as a high availability solution. This is because the replication is asynchronous, which can result in data loss during failures, and it lacks automatic failover and failback capabilities.

Conclusion: Why SafeKit's SAN-Less HA is a Modern Alternative

The core difference lies in the reliance on shared storage. SafeKit's approach uses synchronous data replication directly between servers, ensuring a zero data loss (RPO=0) scenario and dramatically simplifying deployment by eliminating the need for expensive SAN hardware and specialized storage skills. While traditional clusters provide robust features, the high complexity, high cost, and shared-disk dependency make them less flexible, especially for remote or multi-site deployments. For organizations prioritizing cost-effectiveness, simplicity, and fast deployment of VM High Availability without investing in complex storage infrastructure, SafeKit's SAN-less model provides a superior and modern alternative.

NGINX: The Simplest Load Balancing Cluster with Failover

Evidian SafeKit

How the Evidian SafeKit software simply implements NGINX load balancing and failover?

The solution for NGINX

A generic product

A complete solution

Virtual IP address in a farm cluster

Alternative

Alternative

Alternative

Alternative

SafeKit's Advantage in Application Redundancy

Summary and Key Takeaways for High Availability

Summary of SafeKit Mirror Cluster Benefits for High Availability

Summary of SafeKit Farm Cluster Benefits for Load Balancing

Final Recommendation: VM HA for Generality vs. Application HA for Low RTO

Conclusion: Why SafeKit's SAN-Less HA is a Modern Alternative