---
title: "Handling Power Outages and Network Isolation in a HA Cluster"
description: "Handling Power Outages and Network Isolation in a HA Cluster"
canonical: "https://safekit.eviden.com/best-practises/power-outage-and-network-isolation-in-a-cluster/"
topics: "What are the different scenarios in case of network isolation in a cluster?, What are the different scenarios in case of power outage in a cluster?, 🔍 SafeKit High Availability Navigation Hub"
category: "best-practises"
lang: "en"
---

<h1 id="handling-power-outages-and-network-isolation-in-a-ha-cluster">Handling Power Outages and Network Isolation in a HA Cluster</h1>
<h2 id="isolation">What are the different scenarios in case of network isolation in a cluster?</h2>
<h3 id="a-single-network">A single network</h3>
<p>When there is a network isolation, the default behavior is:</p>
<ul>
<li>as heartbeats are lost for each node, each node goes to ALONE and runs the application with its virtual IP address (double execution of the application modifying its local data),</li>
<li>when the isolation is repaired, one ALONE node is forced to stop and to resynchronize its data from the other node,</li>
<li>at the end the cluster is PRIM-SECOND (or SECOND-PRIM according the duplicate virtual IP address detection made by Windows).</li>
</ul>
<h3 id="two-networks-with-a-dedicated-replication-network">Two networks with a dedicated replication network</h3>
<p>When there is a network isolation, the behavior with a dedicated replication network is:</p>
<ul>
<li>a dedicated replication network is implemented on a private network,</li>
<li>heartbeats on the production network are lost (isolated network),</li>
<li>heartbeats on the replication network are working (not isolated network),</li>
<li>the cluster stays in PRIM/SECOND state.</li>
</ul>
<h3 id="a-single-network-and-a-splitbrain-checker">A single network and a splitbrain checker</h3>
<p>When there is a network isolation, the behavior with a split-brain checker is:</p>
<ul>
<li>a split-brain checker has been configured with the IP address of a witness (typically a router),</li>
<li>the split-brain checker operates when a server goes from PRIM to ALONE or from SECOND to ALONE,</li>
<li>in case of network isolation, before going to ALONE, both nodes test the IP address,</li>
<li>the node which can access the IP address goes to ALONE, the other one goes to WAIT,</li>
<li>when the isolation is repaired, the WAIT node resynchronizes its data and becomes SECOND.</li>
</ul>
<p>Note: If the witness is down or disconnected, both nodes go to WAIT and the application is no more running. That&rsquo;s why you must choose a robust witness like a router.</p>
<h2 id="what-are-the-different-scenarios-in-case-of-power-outage-in-a-cluster">What are the different scenarios in case of power outage in a cluster?</h2>
<h3 id="primary-node-power-outage">Primary node power outage</h3>
<p>When a power outage stops only the primary node:</p>
<ul>
<li>there is an automatic failover on the secondary node, which becomes ALONE and restarts the application,</li>
<li>when node 1 is rebooted, it becomes SEDOND after resynchronization of replicated data,</li>
<li>the roles of primary and secondary can be swapped by an adminsitrator if needed.</li>
</ul>
<h3 id="secondary-node-power-outage">Secondary node power outage</h3>
<p>When a power outage stops only the secondary node:</p>
<ul>
<li>there is no failover, the primary becomes ALONE and the application continues its execution on node 1,</li>
<li>when node 2 is rebooted, it becomes SEDOND after resynchronization of replicated data.</li>
</ul>
<h3 id="general-power-outage---case-1">General power outage - case 1</h3>
<p>When a power outage stops both nodes, the default behavior is:</p>
<ul>
<li>both nodes goes to STOP,</li>
<li>when node 1 is rebooted, it does not go into ALONE state and does not restart the application because it doesn&rsquo;t know if it has the up-to-date data. So it goes to the WAIT state waiting for the restart of the other node,</li>
<li>when node 2 is rebooted, both nodes return to their previous PRIM/SECOND states.</li>
</ul>
<h3 id="general-power-outage---case-2">General power outage - case 2</h3>
<p>When a power outage stops both nodes, the behavior with syncdelta is :</p>
<ul>
<li>syncdelta is set for example to 10 minutes in the configuration meaning that start of a node is accepted even if its data is 10 minutes behind the last sync,</li>
<li>when node 1 is rebooted, it goes to ALONE and restarts the application assuming that the restart is done within 10 minutes after the power failure,</li>
<li>when node 2 is rebooted, it becomes SECOND after resynchronization of replicated data,</li>
<li>Note: if node 2 is rebooted the first, then it becomes ALONE and node 1 will become SECOND at its start.</li>
</ul>
<h2 id="hub">🔍 SafeKit High Availability Navigation Hub</h2>
<p><strong>Explore SafeKit: Features, technical videos, documentation, and free trial</strong></p>
<table>
<thead>
<tr>
<th>Resource Type</th>
<th>Description</th>
<th>Direct Link</th>
</tr>
</thead>
<tbody>
<tr>
<td><strong>Key Features</strong></td>
<td>Why Choose SafeKit for Simple and Cost-Effective High Availability?</td>
<td><a href="/#why-choose-safekit-for-ha" title="Discover SafeKit features for simple and cost-effective high availability">See Why Choose SafeKit for High Availability</a></td>
</tr>
<tr>
<td><strong>Use Cases</strong></td>
<td>Explore How SafeKit Ensures the High Availability of Critical Infrastructure</td>
<td><a href="/#safekit-use-cases" title="Explore SafeKit high availability use cases">See All Use Cases (OEM Software, Edge Servers, SCADA, and more)</a></td>
</tr>
<tr>
<td><strong>Deployment Model</strong></td>
<td>All-in-One SANless HA: Shared-Nothing Software Clustering</td>
<td><a href="/#all-in-one-sanless-ha" title="Learn about all-in-one SANless high availability with shared-nothing software clustering">See SafeKit All-in-One SANless HA</a></td>
</tr>
<tr>
<td><strong>HA Strategies</strong></td>
<td>SafeKit: Infrastructure (VM) vs. Application-Level High Availability</td>
<td><a href="/#safekit-ha-redundancy-choices" title="Compare VM-level redundancy with SafeKit application-level high availability strategies">See SafeKit HA &amp; Redundancy: VM vs. Application Level</a></td>
</tr>
<tr>
<td><strong>Technical Specifications</strong></td>
<td>Technical Limitations for SafeKit Clustering</td>
<td><a href="/#safekit-ha-limitations" title="Technical requirements and limitations for SafeKit application clustering">See SafeKit High Availability Limitations</a></td>
</tr>
<tr>
<td><strong>Proof of Concept</strong></td>
<td>SafeKit: High Availability Configuration &amp; Failover Demos</td>
<td><a href="/#safekit-failover-tutorials" title="Step-by-step videos on SafeKit high availability, from installation to automated failover">See SafeKit Failover Tutorials</a></td>
</tr>
<tr>
<td><strong>Architecture</strong></td>
<td>How the SafeKit Mirror Cluster works (Real-Time Replication &amp; Failover)</td>
<td><a href="/#safekit-mirror-cluster" title="See technical architecture and failover mechanism of SafeKit Mirror Cluster">See SafeKit Mirror Cluster: Real-Time Replication &amp; Failover</a></td>
</tr>
<tr>
<td><strong>Architecture</strong></td>
<td>How the SafeKit Farm Cluster works (Network Load Balancing &amp; Failover)</td>
<td><a href="/#safekit-farm-cluster" title="Technical overview of SafeKit Farm Cluster architecture with network load balancing">See SafeKit Farm Cluster: Network Load Balancing &amp; Failover</a></td>
</tr>
<tr>
<td><strong>Competitive Advantages</strong></td>
<td>Comparison: SafeKit vs. Traditional High Availability (HA) Clusters</td>
<td><a href="/#safekit-ha-comparison" title="Detailed comparison of SafeKit software vs traditional hardware-based HA clusters">See SafeKit vs. Traditional HA Cluster Comparison</a></td>
</tr>
<tr>
<td><strong>Technical Resources</strong></td>
<td>SafeKit High Availability: Documentation, Downloads &amp; Trial</td>
<td><a href="/#safekit-ha-technical-resources" title="Access SafeKit free trial, technical documentation, and high availability white papers">See SafeKit HA Free Trial &amp; Technical Documentation</a></td>
</tr>
<tr>
<td><strong>Pre-configured Solutions</strong></td>
<td>SafeKit Application Module Library: Ready-to-Use HA Solutions</td>
<td><a href="/#safekit-ha-application-modules" title="Browse the library of pre-configured SafeKit modules for automated application failover">See SafeKit High Availability Application Modules</a></td>
</tr>
</tbody>
</table>
<div class="button-row">
  <a class="btn-action" href="https://safekit.eviden.com/contact-us-for-safekit/">🧑 Contact us</a>
  <a class="btn-action" href="/resources/safekit-free-trial/">🎁 SafeKit free trial</a>
  <a class="btn-action" href="https://training.my.evidian.com/mod/page/view.php?id=712">🏅 Free certification</a>
  <a class="btn-action" href="https://safekit.eviden.com/get-a-quote-safekit/">💰 Perpetual license cost</a>
</div>