Current Events & Commentary / Product Reviews
The *noMAX HA/DR solution from Maximum Availability ensures unmatched data integrity and optimal recovery time for the IBM System i.
Written by Michael Stuhlreyer
In today's highly competitive global marketplace, it is a business imperative to ensure the continuity of critical business operations and to protect a company's information assets. This makes business continuity an essential objective for companies of all sizes. Businesses must satisfy themselves that they have a comprehensive high availability/disaster recovery (HA/DR) solution in place.
The baseline value of such a solution should be gauged in terms of recovery time and data integrity, which is referred to as the Recovery Time Objective (RTO) and the Recovery Point Objective (RPO). Recovery time is the ability to switch to a backup system with minimal delay when a service interruption or system failure occurs. In doing so, there needs to be a recovery point that makes certain the data on the target machine is exactly the same as the data on the production machine at any given moment.
"What's the point of having an HA/DR solution if you're not role-swap-ready or if it takes days to determine if your systems are in sync and you need to check the integrity of your data?" asks Simon O'Sullivan, Senior VP of Sales for Maximum Availability Ltd. Maximum Availability, developers of the *noMAX suite of HA/DR solutions for the IBM i, has been solving these challenges for a decade while establishing a proven track record with its customers. "You need the confidence of knowing that, regardless of the size and complexity of your IBM i environment, your solution not only delivers true real-time replication and optimal RTO and RPO, but that it is also designed to achieve optimal role-swap performance within the constraints of your environment. We are seeing customer situations whereby they are achieving a role-swap between systems in less than five minutes."
That confidence, according to O'Sullivan, is largely a function of an HA/DR solution's architecture; in the case of the *noMAX suite, the architecture exclusively utilizes the IBM System i's remote journaling technology, which is acknowledged as the most efficient way to move information from one IBM i environment to another.
Prior to remote journaling, he reminds us, HA solutions for OS/400 typically had to send and receive the journal data from within the HA product itself. When IBM introduced remote journaling, this functionality was provided free by the IBM i operating system.
O'Sullivan points out that remote journaling represented a major step forward in terms of the architecture and capability enabling HA on the IBM i platform. It allowed Maximum Availability to create a replication solution uniquely capable of handling the extreme high-volume, real-time replication requirements for the toughest environments while still maintaining the usability and simplicity to make HA/DR available to the broader IBM i marketplace.
Designing for remote journaling, rather than trying to retrofit it later, allowed Maximum Availability to make the technology integral to its solution's code and architecture. The design also provides virtually unlimited scalability, real-time sequential replication with minimum latency, and advanced data protection features, all the while reducing the bandwidth requirements during replication and the system resource demands placed on the production machine.
Another important design feature of *noMAX is that it does not need the system audit journal to detect structural and non-journaled object changes on the production system. HA/DR solutions that require customers to turn on the system audit journal result in extra overhead on the production system.
When objects have been detected as changed via the audit journal and then requested for refresh to the backup system, it is possible for lock conditions to arise, which means the objects cannot be sent at that time. When the objects are saved and sent to the backup system, the object and its following data have to be matched to allow normal replication to continue.
The *noMAX design ensures that the issues surrounding the use of the audit journal have been removed so the replication of object changes can occur in real time and in sync with the data, which then follows. All of this is achieved via remote journaling. No other network-send processes are required.
According to David Ginn, KSB Hospital's Director of Information Systems, a *noMAX customer in Illinois, "Two years ago, it wasn't a big deal if we were down for an hour for backups or three hours for an upgrade or even 12 hours, but with our move to paperless charts, it was mandatory that we have our systems up 24x7. *noMAX provides us with two significant benefits: disaster recovery capabilities and uptime."
Other key features of the *noMAX architecture include advanced data protection and integrity and the use of virtually unlimited concurrent apply threads. The advanced data protection and integrity feature runs continuously. It automatically performs checks on the data as it is being replicated. This ensures the integrity prior to the final update being added to the file on the backup system and verifies that source and target data remain consistent.
The use of multiple concurrent apply threads provides incredible flexibility in maximizing replication performance. It's not uncommon for a company to lose power or network connectivity. When that happens, they'll have a backlog of changes accumulating on the production system that aren't getting replicated to the backup, O'Sullivan explains. "So it's a huge benefit to be able to configure more apply groups. This increases the efficiency and the ability to quickly catch up on the backlog once the power is back on or the network is restored. That's essential because the smaller that window, the less exposure you'll have, and it's all about minimizing exposure. The more restricted you are on the number of apply groups, the longer you're not synchronized. And the longer you're not synchronized, the bigger your backlog is and the more exposed you become. With *noMAX, there is minimum latency, and the locking issues which can arise due to the use of the audit journal are not experienced."
*noMAX is virtually unrestricted from a scalability standpoint, which has been proven in real-world situations. It is run in large-scale telecommunication companies with huge transaction volumes and also in hosting companies with dozens of individual customers.
O'Sullivan emphasizes that the *noMAX suite is not just an enterprise solution or just an SMB solution; it's truly both, scalable across all IBM i environments. In addition, the administration features of the *noMAX suite mean there is no need for a full-time dedicated staff to manage it. The *noMAX suite can be installed and managed remotely and includes an intuitive interface through both the traditional green-screen command input and a GUI format running on a Microsoft Windows PC. This allows customers to easily make changes to their *noMAX systems. By simply dragging and dropping configuration information between systems, setup could not be easier. This has the benefit of keeping the configuration consistent and virtually eliminating human error. The *noMAX user interfaces allow users to quickly view replication status, and any data-synchronization issues are highlighted via integrated alerts.
As KSB Hospital found, the years of experience built into the *noMAX product delivered quick and easy role swaps and true data integrity on the backup machine, which, in the end, is what all HA/DR users strive for.