The objective of this document is to walk you through a scripted Neverfail demonstration. Following this guide will help you effectively demonstrate the core capabilities of the Neverfail product suite to your customers.
Any effective demonstration obviously should be tailored to an individual’s specific needs. If a customer is most interested in application monitoring capabilities, then clearly you should spend more time in the 'Rules' tab of the 'Application' page than in the Configure Server portion of the product. However, as a general rule of thumb, you should loosely follow this guide to ensure the most effective test or demonstration possible.
As you proceed through the demonstration, it is important to reiterate to your customers that Neverfail focuses on six key components in order to achieve high-availability and disaster-recovery:
- Data Protection – Real-time, asynchronous replication capability keeps data safely mirrored on a secondary server, whether in a LAN or a WAN environment.
- Server Protection – Heartbeat messages sent from the passive to the active server ensure that the active machine is up and running at all times. If a failure is detected, the passive server can take over all processing seamlessly and automatically until the source of the failure is identified and rectified. Furthermore, once the Primary server is restored to service, switching back is just as seamless and easy – no downtime (for backups and restores) is required to get back to your original, production state.
- Application Protection – Neverfail Plug-ins provide application-specific service and performance monitoring, to ensure that your protected application(s) are not only running, but running well. Furthermore, Neverfail can take corrective, pre-emptive action to solve problems, without necessarily resorting to a full system failover.
- Network Monitoring – Your active server can monitor the state of both your internal network, as well as its connection to the outside world. If a network failure is identified which might affect the usability of your protected application, failover to your Secondary/passive server can occur automatically. Once again, as soon as the network issue is resolved, switching back is as simple as the click of a button.
- Performance Monitoring – In addition to closely monitoring the performance of your protected application, Neverfail also monitors system performance as well. If your memory, disk, CPU, paging file, etc. show decreases in performance, corrective action can also be taken.
- Site Protection / Disaster Recovery – With out-of-the-box support for WAN deployments, Neverfail also provides protection against catastrophic site failures. All processing can be automatically transferred to your Secondary server at an off-site facility, perhaps at another office or a co-location facility. This guarantees uptime even in the direst of circumstances.
Most importantly, Neverfail is unique in its ability to perform all of these functions without any impact on the end user, providing for a completely seamless failover or switchover operation. Only Neverfail provides this level of protection in such a transparent way.
Testing and Evaluation
Neverfail is the only true high availability solution in the Windows marketplace that keeps end users seamlessly connected to their applications irrespective of the nature of a failure. No other high availability solution monitors every aspect of your environment, including your data, application, network connectivity and server health, and is able to failover with no impact to end users.
To begin your evaluation:
- Launch the Neverfail ‘Manage Server’ interface via the Neverfail start menu group.
- You will now see a summary screen with your server pair listed at the top. This summary screen provides a quick view into the status of all Neverfail server pairs within your environment.
- To manage your server pair, select the appropriate checkbox and click the Open button
Seamless Switchover and Switchback
You are now presented with a Server: Summary of the selected server pair. This summary screen displays the replication status of the system (Replicating), and the current roles of your Primary and Secondary servers (active and passive respectively). It also displays the current synchronization status of both the registry and file system data being protected by Neverfail.
When both the registry and file system are synchronized, you can perform a manual switchover of your applications and workload to the Secondary server quickly and easily.
- Launch a client application (e.g. Microsoft Outlook, SQL Query Analyzer, etc.) and connect to your server, ensuring that it is currently operational.
- Launch a DNS administration console to show the current DNS value for the active server.
- Select the Secondary server and click the Make Active button to transfer the workload gracefully to the Secondary server.
- Test your client application again – this time connected to the Secondary server.
- Refresh your DNS console to show the change made to the DNS record for the active server.
- Select the Primary server and click the Make Active button again to gracefully transfer workload back to the Primary server. (For demonstrations with Microsoft Exchange, you should wait a minimum of 5-10 minutes before performing a switchback, due to inherent delays in Active Directory)
During the switchover process, all of the Windows services related to your protected application(s) will shutdown on the Primary server, in order to ensure that all operations and processes have completed and committed their data to disk. Then, the Neverfail Packet Filter will hide your Primary server from the principal (public) network and subsequently expose the Secondary server. Finally, the same Windows services will be started on the Secondary server, and the switchover process is complete.
One additional step performed in WAN environments is to update the DNS record for the active server to point at the Secondary system. This is necessary in WAN environments since the Primary and Secondary servers will be located in different subnets. In a LAN environment, this is not necessary, as both servers will be given the same IP address.
Note that your client application has remained operational with no need to make any manual configuration changes, or even to shutdown and restart the application. Only Neverfail provides such a seamless switchover, with nearly no impact to end users.
This type of manual switchover is commonly used to perform scheduled maintenance on production systems. Rather than incur a lengthy outage during which end users have no access to their applications, simply switchover to your Secondary server and perform your maintenance on the Primary system, outage free.
Data Replication and Rollback
Neverfail uses a proprietary kernel-mode file-filter driver to intercept file and registry updates made against the active server, and apply those same updates to the passive server. Which updates are actually applied are defined through a set of filters.
Upon installation of a Neverfail Plug-in, appropriate filters are automatically detected and created without the need to perform any manual customization. Filters are also regularly updated by plug-ins during routine operations to ensure that they remain accurate (for example if a new SQL Server database is created in a new file group on the active server, Neverfail would detect this change and self-adjust its replication filters accordingly).
These filters can be viewed and modified on the Data page.
- Select the Data tab and then select the File Filters tab to view your current file filters.
- Click the Rollback button to view a list of existing snapshots. You can manually create a snapshot on either the active or passive server at any time by clicking the Create Shadow button.
- Select the 'Configuration' tab to view and configure automatic snapshot creation. Automatic snapshots are always created on the passive server, but a daily automatic snapshot can optionally be created on the active server.
The file and registry replication process sends updates across the Neverfail channel in near real-time, ensuring that both of your servers remain continuously synchronized, and protecting you from data loss in the event of a failure. Keep in mind that this replication process is subject to any bandwidth constraints that you may have between your servers – the greater the available bandwidth, the more quickly that updates will appear on the passive server.
There is one caveat to performing file and registry replication in near real-time – if any type of corruption occurs to a protected file or registry value on the active server, that corruption will also be replicated to the passive server resulting in two corrupted servers.
In order to protect your servers against potential data corruption, Neverfail employs a Data Rollback Module, which allows for rollback of your entire application state (including data and registry) to a pre-corrupted point in time. Snapshots can be taken at regular intervals, typically on the passive server, and then utilized to recover from any type of corruption event.
Up to 512 separate snapshots can be stored at a time, and the various granularity settings for snapshot frequency and retention offer the most efficient use of storage space.
Furthermore, by automatically taking snapshots of both protected file data as well as registry information, only Neverfail ensures that your entire application state will be consistent following a rollback operation – other products that only allow you to rewind file data risk putting your files and registry in an inconsistent / inoperable state.
The next critical component required to ensure that end users remain productively connected to a functional application is monitoring the server hardware itself for critical failures. Neverfail accomplishes this through the use of a ‘Heartbeat’ message sent across the Neverfail Channel at regular intervals. If a configured number of heartbeats are missed, an automatic failover to the passive server can optionally be triggered.
- Click the Communication button and then select the 'Configuration' tab.
- Verify the ‘Channel Heartbeat’ configuration settings and make any changes as desired. Click Apply after making any changes.
After the configured number of heartbeat messages have been missed, Neverfail will automatically perform a failover operation to the passive server.
After the source of the server problem has been addressed, you must manually restart replication. Neverfail will now begin an automatic resynchronization process between the servers, comparing file and registry data, in order to bring the servers back in sync. While this resynchronization process is underway, the 'Switchover' button is temporarily disabled. Once the servers are synchronized, you can perform a switchback to return application processing to the original server (as the server issue has now been resolved).
Protecting against data loss, data corruption and application problems is useless if users cannot connect to the application itself. If any network problems exist that might make the server seem to be unavailable, including issues with network hardware, cabling, or routing, end users will perceive a server outage even when one does not exist.
In order to ensure that a server can see other locations around the network, Neverfail offers the ability to ping various IP addresses and to switchover to the passive server if none of the pinged addresses respond.
- Click the Communication button and then select the 'Configuration' tab.
- Select the ‘Failover If Active Server’s Public Network Connection Is Lost’ check box and then click the Apply button.
After the configured number of ping messages has been missed, Neverfail will automatically perform a graceful switchover to the passive server.
As a precautionary measure, Neverfail also turns off replication in the opposite direction, consequently leaving the server pair ‘Out of Synch’. After the source of the networking problem has been addressed, you must manually restart replication. Neverfail will now begin an automatic resynchronization process between the servers, comparing file and registry data, in order to bring the servers back in sync. While this resynchronization process is underway, the 'Switchover' button is temporarily disabled. Once the servers are synchronized, you can perform a switchback to return application processing to the original server (as the networking issue has now been resolved).
Upon completion of the switchback, all of your end users are once again connected to the Primary / active server without any need to restart their client applications.
Neverfail is the only high availability application that offers the ability to ping multiple IP addresses both inside as well as outside of your network in order to ensure connectivity. Network connectivity monitoring and automated switchover is a critical component to ensure that end users remain seamlessly connected to their applications.
Application and Performance Monitoring
The ability to replicate data to a passive server may protect you against data loss and rollback capabilities may help you recover from data corruption, but what about the applications that actually make use of that data?
It is important to ensure that your protected applications are actually operational and providing service to your end users, and that the performance of those applications is adequate – otherwise, your end users are not being productive. Neverfail Plug-ins provide these monitoring and pre-emptive repair capabilities.
- Connect to the active Neverfail server by double-clicking the appropriate desktop icon.
- Once you’ve connected to the active server (via RDP connection), launch the Neverfail Heartbeat Management Client on the active server via the Neverfail Start menu group.
- Click on the Application button and select the Tasks tab. Select the 'Service Monitor' task from the list of tasks provided, click on the Edit button, and set the Interval to a low value. Click OK .
Click on the Services tab. Select a service from the list and click on the Edit button. Set the ‘First Failure’ option to ‘Restart Applications’. Click Ok after making any changes.
Use the Windows service control manager to stop the selected service manually. Wait for the sampling frequency to pass and then refresh the service control manager to verify that the service was successfully restarted.
Select the Rules tab to view and configure application monitoring rules.
Depending upon the application(s) that you are protecting, under the 'Services' tab you will be provided with a list of appropriate Windows services that you can monitor and corrective actions that can be taken whenever a service has failed. Using the 'Tasks' tab, you can also modify the sampling frequency that Neverfail uses to check the services.
In addition to monitoring application services, Neverfail can also monitor specific application attributes to ensure that they remain within normal operating ranges. Similar to application monitoring, various rules can be set to trigger specific corrective actions whenever these attributes fall outside of their respective ranges.
Only Neverfail monitors application services as well as specific application attributes to ensure that your protected applications are operational and not in a ‘hung’ or ‘stopped’ state. This level of monitoring is fundamental in ensuring that your applications are actually available to end users.
Furthermore, no other high availability application provides the same level of flexibility in defining multiple corrective actions to be taken in the event of problems; on a service-by-service or even attribute-by-attribute basis.
The ultimate goal of high availability and disaster recovery is to keep end users seamlessly connected to their applications irrespective of the nature of a failure. In order to achieve this goal successfully, a high availability product must provide the following critical capabilities:
Seamless Switchover and Switchback – end users continue using their applications with minimal impact, requiring no client application configuration changes or restarts.
Data Replication and Rollback – file and registry replication keeps data safely mirrored on a passive server, whether in a LAN or a WAN environment, and data rollback capabilities protect against data corruption by rolling back the entire application state if necessary.
Server Monitoring – if a critical hardware failure is detected, a passive server can take over all processing seamlessly and automatically until the source of the failure is identified and rectified.
Network Monitoring – if a network failure is identified which might affect the usability of your protected application, automatically switchover to your passive server.
Application and Performance Monitoring – application-specific service and performance monitoring ensures that your protected application(s) are not only running, but running well.
Any product, which does not address all of the above requirements, is not a complete solution.