SIOS SANless clusters

SIOS SANless clusters High-availability Machine Learning monitoring

  • Home
  • Products
    • SIOS DataKeeper for Windows
    • SIOS Protection Suite for Linux
  • News and Events
  • Clustering Simplified
  • Success Stories
  • Contact Us
  • English
  • 中文 (中国)
  • 中文 (台灣)
  • 한국어
  • Bahasa Indonesia
  • ไทย

Fail Open vs. Fail Closed: Why It’s a False Dilemma (and How True HA Fixes It)

November 14, 2026 by Jason Aw Leave a Comment

Fail Open vs Fail Closed Why Its a False Dilemma (and How True HA Fixes It)

Fail Open vs. Fail Closed: Why It’s a False Dilemma (and How True HA Fixes It)

Every infrastructure architect eventually hits the wall of inline security design: If our security stack crashes, do we prioritize business continuity or perimeter defense?

This decision surfaces across Web Application Firewalls (WAFs), email hygiene gateways, Next-Gen Firewalls (NGFWs), and API proxies. It is often framed as an unavoidable trade-off between two undesirable outcomes: Fail Open or Fail Closed.

What Is Fail Open vs. Fail Closed?

When an inline inspection node suffers an unhandled exception, resource exhaustion, or kernel panic, it must resolve network traffic in one of two ways:

  • Fail Open (Bypass Mode): Network traffic flows directly through to the destination without inspection.
    • The Upside: Users don’t experience an outage. Business transactions proceed, latency stays low, and uptime metrics stay green.
    • The Downside: The perimeter vanishes. Adversaries can intentionally crash an engine (via algorithmic complexity attacks, malformed packet bursts, or memory leaks) specifically to force it into bypass and deliver payloads undetected.
  • Fail Closed (Drop Mode): All traffic routed through the failed inspection appliance is dropped.
    • The Upside: Security integrity remains intact. No uninspected, potentially malicious byte reaches internal systems.
    • The Downside: Total denial of service. For e-commerce checkout APIs, authentication gateways, or customer-facing portals, this means immediate revenue loss and angry customers.

How This Plays Out Across Security Appliances

Appliance Type Fail Open Behavior Fail Closed Behavior The Real-World Risk
Email Spam / Mail Security Gateways Bypasses AV/phishing filters; sends raw inbound mail straight to exchange/inbox. Refuses SMTP connections (4xx temporary error or 5xx reject). In Fail Open, a wave of ransomware or spear-phishing hits inboxes. In Fail Closed, business communication freezes, and external senders get bounce-backs.
Web Application Firewalls (WAFs) Reverts to a plain layer-4/7 reverse proxy, passing SQLi, XSS, and bot traffic through. Returns 502 Bad Gateway / 504 Gateway Timeout or drops TCP sessions. In Fail Open, web assets are completely exposed to zero-days. In Fail Closed, customer traffic stops entirely.
DDoS Scrubbing / IPS Nodes Hard bypass via optical/bypass switches; passes unfiltered traffic upstream. Interface shuts down; link goes dark. In Fail Open, transit pipes saturate immediately. In Fail Closed, the service goes offline—achieving the attacker’s goal anyway.
Identity & Authentication Proxies Grants fallback access or drops MFA enforcement. Rejects all logins (Access Denied). In Fail Open, unauthorized access spikes. In Fail Closed, zero staff or customers can authenticate.

Why “Fail Open vs. Fail Closed” Is an Architecture Antipattern

Relying on an appliance’s native fallback mechanism assumes that failure must mean compromise or downtime. It accepts single points of failure at the architectural level.

A bypass switch or a static drop rule only addresses what to do after an instance dies. It does nothing to:

  1. Re-establish active inspection immediately.
  2. Maintain synchronization of active connection states or security session tables.
  3. Automatically remediate the underlying service fault.

Choosing between security and uptime is usually a sign that the appliance lacks true Application-Aware High Availability (HA).

Escaping the Binary: How something like SIOS LifeKeeper Solves the Dilemma

Rather than deciding whether your crashed WAF, proxy, or MTA should blindly drop or blindly bypass traffic, SIOS LifeKeeper eliminates the dead-end entirely by keeping the inspection layer highly available.

1. Deep Application-Aware Health Monitoring

Standard hypervisors or cloud orchestrators only look at whether a virtual machine is running. If the underlying Linux OS is alive but the spam filter’s amavisd, clamav, or custom WAF inspection worker hangs in a deadlock, hypervisors see a healthy VM.

SIOS LifeKeeper includes Application Recovery Kits (ARKs) for many applications but also includes the ability to make your own custom “GenApp ARK” to provide customized monitoring that query the application layer directly:

  • Is the filtering daemon responding to probe requests?
  • Are memory limits approaching threshold panics?
  • Are the network sockets actively processing queues?

2. Local Self-Healing First, Failover Second

Before triggering a costly cross-node failover, LifeKeeper can attempt an automated, controlled restart of the misbehaving security service. If the process is irrecoverable, it orchestrates an automated failover.

3. Zero-Data-Loss Configuration & State Replication

Security appliances often depend on dynamic blocklists, rate-limiting caches, and session states. In tandem with SIOS DataKeeper, LifeKeeper provides block-level synchronous or asynchronous storage replication without requiring expensive, shared SAN hardware. The standby node has access to the exact policy sets, certificates, and IP reputation caches the moment it assumes control.

4. Automated Virtual IP and Network Redirection

When failover occurs, LifeKeeper reallocates the cluster Virtual IP (VIP), updates cloud routing tables (e.g., across AWS AZs, Azure Availability Zones, or on-premises subnets), and starts the security services on the secondary node.

The result:

  • The system behaves like “Fail Closed” from a security perspective: uninspected packets are never let through blindly.
  • The system behaves like “Fail Open” from an uptime perspective: the service recovers automatically in seconds, eliminating prolonged outages.

Real-world example… Protecting ModSecurity with SIOS LifeKeeper

As a widely adopted open-source web application firewall engine embedded within Apache HTTP Server, Nginx, or IIS, ModSecurity sits directly inline to evaluate incoming HTTP traffic against malicious signatures. Consequently, when an unhandled daemon lockup or process crash occurs, infrastructure teams face a stark operational dilemma: expose the application to immediate compromise or accept a total service outage. This makes ModSecurity an ideal scenario for leveraging SIOS LifeKeeper to resolve the trade-off through robust High Availability.

Custom GenApp ARK Monitoring for Web Server Processes

Using the SIOS LifeKeeper Generic Application Recovery Kit (GenApp ARK), administrators can define custom health-check scripts tailored to ModSecurity and its host web server (Apache or Nginx):

  • Regularly probe web server worker processes and verify that the ModSecurity engine is actively parsing requests without deadlocks.
  • Monitor rule engine memory usage and HTTP response codes to detect silent inspection failures.

Local Service Self-Healing

If the GenApp ARK detects that the Apache or Nginx process hosting ModSecurity has hung or crashed, LifeKeeper first attempts an automated local restart of the web server and WAF module. This resolves transient software faults immediately without interrupting cluster state or switching node ownership.

Automated Failover and Rule Synchronisation

If local recovery fails, LifeKeeper automatically initiates failover to a standby node. In conjunction with SIOS DataKeeper block-level replication, custom SecRule directives, audit logs, and IP reputation tables are continuously synchronised, ensuring the secondary node assumes active WAF inspection with up-to-date protection profiles.

The Verdict

“Fail open or fail closed” is a forced choice that stems from treating security nodes as isolated, standalone appliances.

For perimeter security where data breaches and downtime carry identical financial risk, the goal should be neither. By anchoring your email security clusters, WAFs, and proprietary Linux/Windows security servers to an active-passive or active-active framework via SIOS LifeKeeper, you take the human gamble out of the equation: security stays rigid, and operations stay online.

About the Author: Aaron West is a Solutions Engineer at SIOS Technology Corp, specializing in high availability, open-source solutions, and networking. With over a decade of experience, Aaron has held leadership roles, including Head of Solutions at Loadbalancer.org, where he built and led teams in technical support and pre-sales consulting. His expertise spans network engineering, Linux, and Microsoft systems, with a strong focus on team building and customer-centric solutions. Aaron is passionate about delivering robust open-source technologies and providing tailored solutions for clients across various industries.

Reproduced with permission from SIOS

Filed Under: Clustering Simplified Tagged With: cluster, High Availability

The 5-Step Checklist for Validating Your Failover Setup

November 8, 2026 by Jason Aw Leave a Comment

The 5-Step Checklist for Validating Your Failover Setup

The 5-Step Checklist for Validating Your Failover Setup

Having a failover environment configured is only one part of completing the setup. Validating and verifying whether your configuration will work as expected is the next and most important step. Here are ways you can be confident your infrastructure can withstand a disaster.

Build a QA Cluster That Mirrors Production

A QA cluster is typically a replica of the production cluster that is used for testing and/or implementing any changes that need to be introduced into the cluster. With environments that contain multiple failover solutions, it’s necessary to have an environment that allows for testing to see how the environment behaves. When solutions work separately, there is sometimes an assumption that they will seamlessly work together. This may not always be the case. Finding out that solutions do not work well together could happen during a failover. If the order of operations is not tested and verified to work, you may experience a failure rather than a failover, possibly due to the order of operations. Having a QA environment is useful in mitigating and preventing issues arising when HA/DR response is needed most.

Test Failover and Switchover Scenarios

Testing is crucial in validating your failover setup. Testing helps ensure that an environment behaves properly when a failure scenario happens. Simultaneously, testing is important to verify if the behavior of the environment aligns with your expectations. There are circumstances where you may not want to begin a failover immediately, which is why it helps to understand the difference between switchover and failover testing.

Verify Logs Before Go-Live

Before going into production, always check the logs from your system and the failover solution, such as SIOS LifeKeeper for Linux. Ideally, you will want to verify that within your environment all needed systems, clusters, and infrastructure elements can communicate with each other to ensure smooth operation. Checking the logs for any anomalies is also helpful in understanding how your environment is behaving, especially during testing, should your expectations not be met.

Document Everything in a Runbook

Always have a Runbook to document procedures for the environment and to outline the tests that should be executed. Documenting the process of setting up the environment and testing the environment helps beyond having an understanding of what you are doing and the information that is being shared. Documentation makes sure processes are repeatable, but it also serves as a record that you can audit if needed. By keeping documentation of how procedures should be performed and the tests that should be verified, a paper trail can be maintained so any analyses performed after an issue opens up room for improvement.

During this phase of validating your configuration, it is not expected that everything will be 100% successful; failures are inevitable. When they do happen, a runbook provides documentation on the steps that led to the situation. Runbooks and documentation are important because they provide insight into the events that did – or did not – happen in the lead-up to an issue. With thorough documentation, you get the ability to adjust procedures and prevent these circumstances from creating an issue in your production environment.

Confirm Application Setup Meets Your Failover Solution’s Requirements

Certain failover solutions may differ in their requirements for application protection. Setting up the applications is sometimes not as simple as making sure the application itself works, but making sure the application and the failover solution work well together. Verify with the respective documentation on how applications need to be set up.

For example, the LifeKeeper for Linux documentation outlines a requirement to modify some default settings to allow proper protection of an SAP environment, as seen in the modify ASCS and ERS instance profile settings for SAP page.

Considerations like these are crucial to ensure that your HA/DR solution and protected applications work together cohesively and grant the most value from the solutions in use.

When deploying an environment, there are a lot of factors at play. With this checklist, rollouts can be much more efficient and manageable. This checklist will help ensure your environment is validated and ready for production, giving you confidence in the life of your deployment. Want an expert to check your setup? Schedule an HA Health Check with SIOS.

About the Author: Alexus Gore is a Customer Experience Software Engineer at SIOS Technology Corp. She focuses on helping customers optimize their high availability and disaster recovery environments, ensuring their systems are resilient and reliable.

Reproduced with permission from SIOS

Filed Under: Clustering Simplified Tagged With: cluster, failover

Managing an Inherited DataKeeper Cluster: Challenges and Best Practices

October 31, 2026 by Jason Aw Leave a Comment

Setting Up Nginx Load Balancer on AWS

Managing an Inherited DataKeeper Cluster: Challenges and Best Practices

How’s that inheritance coming along? Part 2 of our Inheriting DataKeeper series.

Inheriting an IT environment is rarely a one-and-done process, and inheriting a DataKeeper-enabled cluster is no exception. When we first discussed the idea of “inheritance” in the context of DataKeeper in our first post, Inheriting DataKeeper,, we emphasized the importance of getting your bearings—understanding the environment, asking the right questions, and collaborating across multiple teams. But once the “honeymoon” phase is over, the real work begins. So, how’s it going? Are you thriving in your new role as steward of your inherited cluster, or are you slowly discovering the quirks, gaps, and potential pitfalls lurking beneath the surface?

In this follow-up blog, let’s revisit some of the concepts from the original post, dig deeper into the challenges an administrator may now be facing, and explore strategies for ongoing success as you continue to manage your DataKeeper inheritance.

Key Takeaways

  • Treat your DataKeeper runbook as a living document and update it after every change or test.
  • Map the dependencies between DataKeeper mirrors, SQL Server, storage, and networking—a change in one can ripple through the others.
  • Schedule failover tests regularly; an untested HA/DR configuration is an assumption, not a plan.
  • Track licensing renewals, release notes, and patches to avoid surprise service interruptions.
  • Use monitoring and alerts to catch replication delays before they impact workloads.

Reality Check: What You’ve Learned About Your Inherited Cluster

If you’re a few weeks or months into your role as the inheritor of a DataKeeper environment, you’ve likely gained some valuable insights—both about the technology and about the infrastructure-specific nuances of your cluster. Here are some common “aha moments” I’ve encountered in and throughout my career:

1. Runbooks Are Great—Until They’re Not
Remember how we stressed the importance of a well-documented runbook? By now, you’ve probably realized that runbooks can be hit-or-miss. If your predecessor was meticulous in their documentation, congratulations—you’re in great shape. But if you’ve inherited a runbook that’s outdated, incomplete, or nonexistent, you may be feeling the strain of piecing together critical workflows on your own.
Personal Tip: Make it a priority to update your runbook as you uncover gaps or introduce changes. Treat it as a living document, not a dusty binder in the corner.

2. Dependencies Are Sneaky
IT systems are like ecosystems—everything is interconnected. You may have discovered that changes to one component (e.g., SQL Server configurations, storage volumes, or network settings) can ripple through your cluster in unexpected ways. Identifying these dependencies is an ongoing process, and it requires constant communication with cross-functional teams.
Personal Tip: Build strong relationships with the SQL Server, Networking, and Storage teams, and establish regular touchpoints to share updates and flag potential issues early.

3. Failover Testing Is Not Optional
If you’ve already performed a failover test (or several), kudos to you. If you haven’t—what are you waiting for? Failover testing is critical for ensuring your high availability (HA) and disaster recovery (DR) configurations work as intended. Just because everything looks good on paper doesn’t mean it will behave as expected during an actual failover scenario.
Personal Tip: Schedule failover tests regularly, and involve key stakeholders in the process. Document the results, and update your runbook accordingly.

Common Challenges in the “Inheritance Phase”

Managing your inherited cluster is an ongoing process, and it’s not without its challenges. Here are a few common pain points that may be surfacing as you settle into your role:

1. Licensing and Account Management
Are you still tracking licensing renewals and account details? If not, you could be on a collision course with unexpected service interruptions. Stay vigilant about renewal deadlines, and ensure your credentials for the licensing portal are up to date.

2. Performance Bottlenecks
As workloads evolve, your cluster’s performance may be impacted. Have you noticed increased latency or replication delays? It may be time to revisit your storage bandwidth, mirror types, or network configurations, just to name a few.
Personal Tip: Collaborate with the Networking and Storage teams to identify and address bottlenecks before they escalate.

3. Software Updates and Patch Management

Keeping your DataKeeper environment up to date is essential for security, stability, and performance. But coordinating updates across multiple systems can be tricky. Are you proactively monitoring for new releases and patches?
Personal Tip: Establish a patch management strategy that includes regular updates, testing procedures, and rollback plans.

 4. Knowledge Silos (Yes!  It’s a thing at a Data Center, and that’s where I picked up the lingo)
If your predecessor didn’t do a thorough knowledge transfer—or if key team members have left—critical information about your cluster may be locked away in silos. Bridging these gaps requires patience, detective work, and collaboration.
Personal Tip: Host regular knowledge-sharing sessions with your team to surface hidden insights and ensure everyone is aligned.

Strategies for Longterm DataKeeper Cluster Health

So, what’s next? How do you ensure your inherited DataKeeper environment continues to run smoothly while avoiding burnout or unexpected surprises? Here are a few strategies to keep in mind:

1. Proactive Monitoring and Alerts
Set up monitoring tools and alerts to keep an eye on your cluster’s health. By proactively identifying issues—such as replication delays or resource contention—you can address them before they impact your workloads.

2. Continuous Learning
DataKeeper is a powerful tool, but it’s also nuanced. Take advantage of training resources, webinars, and documentation provided by SIOS to deepen your understanding of the technology.
Personal Tip: Bookmark the SIOS Resource Center for easy access to guides, whitepapers, and case studies.

3. Cross-Functional Collaboration
The success of your cluster depends on collaboration across teams. Build strong relationships with key stakeholders, and establish regular communication channels to keep everyone in the loop.

4. Prepare for Growth
Is your cluster environment expected to scale in the near future? Whether you’re adding new nodes, workloads, or mirrors, ensure you have a plan in place to support growth without compromising performance or availability.

The Final Question: How’s It Going?

Inheriting a DataKeeper environment is both a privilege and a responsibility. It’s an opportunity to build on the foundation laid by your predecessor while introducing improvements, efficiencies, and innovations of your own. So, how’s it going? Are you embracing the challenge, or are you still finding your footing? Wherever you are in your journey, know that you’re not alone. Lean on your team, your documentation, and resources like SIOS to guide you through the process.

Remember: your inheritance is only as valuable as the care and attention you give it. Keep learning, keep collaborating, and keep improving—and you’ll transform your inherited cluster into a thriving environment that supports your organization’s goals for years to come.

Frequently Asked Questions

What should I do first after inheriting a DataKeeper cluster?

Start by getting your bearings: confirm who owns the licensing and support accounts, review (or create) the runbook, inventory your mirrors and cluster resources, and identify the SQL Server, storage, and networking teams whose changes can affect the cluster. Then schedule a failover test to verify the environment works as documented.

How often should I test failover in a DataKeeper cluster?

Test on a regular schedule and after any significant change to SQL Server, storage, networking, or DataKeeper itself. [Greg: add recommended cadence.] Document the results of each test and update your runbook so everyone knows what was verified and when.

What should a DataKeeper runbook include?

A DataKeeper runbook should list every mirror (source and target volumes, synchronous or asynchronous mode), the DataKeeper service account, license details and renewal dates, WSFC resource dependencies, switchover and failback procedures, failover test history, and escalation contacts for SIOS Support and your internal teams.

What causes replication delays in a DataKeeper cluster?

Replication delays usually trace back to limited network bandwidth or high latency between nodes, storage performance limits, a mirror type that doesn’t suit the connection (for example, synchronous mirroring over a long-distance link), or workloads that have outgrown the original design. Work with your networking and storage teams to pinpoint the bottleneck.

How do I keep DataKeeper up to date safely?

Monitor SIOS release notes, confirm compatibility with your Windows Server and SQL Server versions, test updates outside production where possible, and keep a rollback plan. Follow the documented DataKeeper Cluster Edition upgrade procedure, then update your runbook with the new version and any configuration changes.

Author: Greg Tucker, Senior Product Support Engineer at SIOS

Filed Under: Clustering Simplified Tagged With: cluster

How to get the most out of your “GET” commands in DataKeeper

September 5, 2024 by Jason Aw Leave a Comment

How to get the most out of your “GET” commands in DataKeeper

How to get the most out of your “GET” commands in DataKeeper

In part 2 of the three-part DataKeeper dashboard blog series, this blog is a follow up from DataKeeper UI vs. Car Dashboards blog. As with your car, when those indicators (traffic light colors) on your dashboard flash, you pop the hood to identify what they represent. Starting points:

  • Battery Light = check battery, cable connections, alternator, etc.
  • Oil Light = checking the dipstick for low, high or no oil (No oil, DON’T MOVE YOUR CAR)
  • Coolant Light = Is there any coolant/water in the overfill tank?

“Popping the hood” on your DataKeeper Cluster Edition software has several similarities and it often means using command line interface for DataKeeper Administration. As for Using EMCMD (Extended Mirroring Command) with SIOS DataKeeper, the GET commands make up 1/3 of the most commonly used tasks/commands of the approximately 48 commands. Note that they are informational only and will not impact your nodes. Below are a few helpful Get commands and their usage to identify the reasons for warning colors in your DataKeeper User Interface/DataKeeper.msc (traffic light colors) in the areas of Storage, Networking and Other.

Note:

From an elevated/Administrator command prompt as always use:

cd %extmirrbase% (This is just a shortcut to the install path <root>\Program Files (x86)\SIOS\DataKeeper>)

Obtaining Status of your DataKeeper Mirrors

A great place to start your initial triage below:

  • getmirrorvolinfo (Other) It is the most heavily used tool as it provides the Mirror roles, Source/Target and the 5 different states a Mirror can be in. It can be run on the source or the target to check to see if the mirror configuration exists.
  • getserviceinfo (Other) – Great information about
    • the Driver Version  and DataKeeper Service Version
    • and the time the SIOS DataKeeper Service was started/restarted
  • getcompletevolumelist (Storage/Network) – a total list
    • of all disks
    •  DataKeeper and non-DataKeeper Volumes
    •  their Role (Source/Target) and Total Capacity (in bytes)
  • getjobinfo (Storage/Network) this shows/echos the job information listed in the DataKeeper console
  • getvolumeinfo (Storage,Network) Great utility when
    • comparing the Total Drive Capacity of the Source (diskmgmt.msc) and Target Volumes (bytes)
    • IP addresses and present Mirror status

Cross-referencing DataKeeper “GET” commands to various Windows Server Commands

  • Getjobinfo and “ipconfig /all” (Networking)
    • Are the IP addresses identical in both outputs?  If the Mirror Replication is to be segmented to another network then the getjobinfo must match the ipconfig /all output.  Executing a changemirrorendpoints will correct this discrepancy (to be discussed in another blog)
      • Also the mirrorendpoints information from the getjobinfo output should match the IP addresses locate in the ipconfig /all output
  • After performing a Resize of the existing Source and Target Volumes, does the emcmd . getvolumeinfo command reflect a different Total Space than the DISKPART command and its Size? Via DISKPART a “extend” or “extend filesystem” may be required so as the resized volumes are recognized by the Operating System
  • Getcompletevolumelist has information similar to that of the Disk Management applet (diskmgmt.msc)
    • File System type
    • Volume Total Space (bytes)
    • Getvolumeinfo provides similar output
  • Getserviceinfo as well as “net start” can provide status about the SIOS DataKeeper Service; if it is Start/Stop

Take Charge of DataKeeper: Apply Your GET Command Knowledge Now

Now that you are armed with some basic knowledge about the lights on your car’s dashboard you will become a DIYer when it comes to DataKeeper Administration in blog 3 in the DataKeeper Dashboard series.

Reproduced with permission from SIOS

Filed Under: Clustering Simplified Tagged With: cluster

How to Set Up a DataKeeper Cluster Edition (DKCE Cluster)

January 30, 2024 by Jason Aw Leave a Comment

How to Set Up a DataKeeper Cluster Edition (DKCE Cluster)

How to Set Up a DataKeeper Cluster Edition (DKCE Cluster)

What is a DKCE cluster?

DKCE is an acronym for DataKeeper Cluster Edition. DKCE is a SIOS software that combines the use of DataKeeper with features of Windows Failover Clustering to provide high availability through migration-based data replication.

The steps to create a DKCE cluster

For this example, I will set up a three-node cluster with the third node maintaining node majority.

Step 1: You must have DataKeeper installed on 2/3 of your systems to set up a DKCE cluster. Click on the following link to follow our quick start guide to complete this install: https://docs.us.sios.com/dkce/8.10.0/en/topic/datakeeper-cluster-edition-quick-start-guide

Step 2: Add the servers that you plan on managing in the Server Manager. This will need to be done on all servers you plan on adding to the cluster.

On your server navigate to Server Manager

Click “Add other servers to manage”

I added the servers by their name here. To do so this way you will need to verify your system name and IP entries in the host file, located here: C:\Windows\System32\drivers\etc\hosts

After all servers have been added, you can verify by navigating to “All servers” in Server Manager

Step 3: You may notice a winRM error, to bypass it run this command in PS as an administrator. Run this command to add servers in your cluster as trusted hosts. This command will need to be run on every system in your cluster.

Set-Item WSMan:\localhost\Client\TrustedHosts -Value ‘<name of server 1>,<name of server 2>’

Step 4: Install Failover Clustering

Follow these steps to install failover clustering.

Step 5: Navigate to the Failover Cluster Manager

Step 6: Click “Create Cluster”

Step 7: Next, add the servers that should be in the cluster and click “Add” after each entry.

Step 8: The list should be similar to the one in the following image

Step 9: Choose “Run all tests” for the validation test, and click Next.

Step 10: Once the tests have been completed, click “Finish”

Step 11: Name your cluster, I have named this one “Cluster1”, click “Next”

Step 12: Verify “Add all eligible storage to the cluster” is checked, click “Next”

Step 13: Once step 12 is completed, click “Finish”

Step 14: In Failover Cluster Manager, the cluster will initially be offline. I will be assigning an unused IP to bring it online. In the “Cluster Core Resources” right-click on the IP address resource and select Properties.

In the properties panel, my subnet mask is /28, therefore I will choose an available IP within the range, 12.0.0.14. Click “Apply”

In the “Cluster Core Resources” right-click on the cluster and select “Bring Online”

The resources should now be online

Step 15: Navigate to DataKeeper

Step 16: Right-click on Job and click “Create Job” to begin creating our first mirror

Give the job a name I am naming my job “job1”, and click “Create Job”

Choose the source and volume to replicate data from. I have chosen Box1 as my source, and volume D, click “Next”

Next, choose a server and a volume to be the target. I have chosen Box2, and volume D.

A prompt will appear to ask you to auto-register the volume you have created as a WSFC volume,  select “Yes” to make this volume Highly Available.

In DataKeeper you can now see that the volume is currently mirroring.

Step 17: In Failover Cluster Manager, navigate to Storage, then Disks. You will see the volume that you have auto-registered is WSFC.

Step 18: Let’s verify the Owners that should be checked. Right-click on the volume, and click “Properties”

Since I need a third to be a witness and maintain node majority, Box3 will need to be unchecked / remain unchecked.

Step 19: Now we can test a migration through Failover Cluster Manager. Navigate to File Explorer and create a new text file in the volume that is currently being mirrored. Do this on your Source.

Navigate to Failover Cluster Manager, click “DataKeeper Volume D” and select “Move Available Storage” from the Actions pane.

Right-click “Best Possible Node”. This should automatically migrate to your target.

In Failover Cluster Manager verify the owner of “DataKeeper Volume D” is now the target node

Navigate to DataKeeper to verify that your target is now the Source, and vice versa.

Successful DKCE Cluster Setup

You have completed the setup of a DKCE cluster.

SIOS provides resources and training for all our products.

Reproduced with permission from SIOS

Filed Under: Clustering Simplified Tagged With: cluster, DKCE

  • 1
  • 2
  • 3
  • …
  • 9
  • Next Page »

Recent Posts

  • Fail Open vs. Fail Closed: Why It’s a False Dilemma (and How True HA Fixes It)
  • The 5-Step Checklist for Validating Your Failover Setup
  • Managing an Inherited DataKeeper Cluster: Challenges and Best Practices
  • Setting Up Nginx Load Balancer on AWS
  • With Great Power: Cluster Maintenance Best Practices

Most Popular Posts

Maximise replication performance for Linux Clustering with Fusion-io
Failover Clustering with VMware High Availability
create A 2-Node MySQL Cluster Without Shared Storage
create A 2-Node MySQL Cluster Without Shared Storage
SAP for High Availability Solutions For Linux
Bandwidth To Support Real-Time Replication
The Availability Equation – High Availability Solutions.jpg
Choosing Platforms To Replicate Data - Host-Based Or Storage-Based?
Guide To Connect To An iSCSI Target Using Open-iSCSI Initiator Software
Best Practices to Eliminate SPoF In Cluster Architecture
Step-By-Step How To Configure A Linux Failover Cluster In Microsoft Azure IaaS Without Shared Storage azure sanless
Take Action Before SQL Server 20082008 R2 Support Expires
How To Cluster MaxDB On Windows In The Cloud

Join Our Mailing List

Copyright © 2026 · Enterprise Pro Theme on Genesis Framework · WordPress · Log in