Shield Launches ‘DR for the Masses’
April 17, 2007 Alex Woodie
Shield Advanced Solutions last week unveiled a new disaster recovery (DR) solution for the i5/OS server, called Receiver Apply Program/400 (RAP/400). The new product is based on the remote journaling technology IBM developed years ago, primarily for ensuring high availability (HA) of AS/400s. But instead of high availability, Shield is using remote journaling as a way to bring high-speed data replication and next-generation DR to the i5/OS masses.
There is no question that remote journaling has had a very large impact on IBM’s AS/400-iSeries-System i business over the last seven years. Before remote journaling was available, providers of high availability software had to develop their own data replication technologies, which involved “scraping” the local journal for changes to data and objects, and then ferrying the changes across a network to the target machine. With remote journaling embedded below the operating system layer starting with OS/400 V4R4, much of the development work for the HA data transport mechanism–and many (but not all) of the problems high availability software vendors deal with–were eliminated.
Due to the legacy code base, it took a while for remote journaling to take hold and gain market acceptance. Some of the big high availability vendors were slow to sing to the praises of remote journaling, in part because it dropped the technical barrier of entry into a lucrative business, and they didn’t want to give credibility to newcomers building on the remote journaling technology, such as iTera, which has since merged with Vision Solutions. Even Shield Advanced Solutions, which has partnerships with the big HA providers, at first questioned the efficacy of remote journaling, although the company soon after saw the light and became a remote journaling convert.
iTera’s market success proved remote journaling was no fluke, and eventually even the big guys, such as Vision and Lakeview Technology, started selling more high availability software that uses remote journaling. Today, nearly all new high availability implementations are utilizing remote journaling.
Now, with RAP/400, Shield is hoping to take the basic data transport mechanism enabled by remote journaling and use it for high availability’s little brother: disaster recovery.
With RAP/400, customers gain a way to apply changes that arrive in a journal receiver attached to a remote journal. RAP/400 automatically applies all the entries from the journal every time the receiver is detached from the target journal. These changes can be applied to objects stored on the backup iSeries or System i machine, or in a secondary LPAR of the primary machine. If RAP/400 is unable to apply changes to any object, the event gets logged in an output file, and customers will have to manually re-sync the objects.
Executing receiver applies is a feature that RAP/400 has in common with all high availability products. After all, remote journaling just ensures that the changes arrive at their destination–it’s up to the customer or HA product to do something with the changes once they get there. However, RAP/400 provides this apply functionality at a greatly reduced price compared to a full blown HA solution. It also lacks many other features common to HA–most notably, a failover mechanism–and many other HA bells and whistles.
RAP/400 will support a single journal receiver, and will support only those objects that are supported by remote journaling. The product includes APIs and requires i5/OS V5R4.
Shield says RAP/400 won’t compete with high availability products. (Shield is a participant in the industry with its tool, JobQGenie, which ensures continuity of scheduled jobs in high availability environments.) Instead, the main RAP/400 users will be customers looking to improve upon their DR strategies, which today are mostly tape-based, as well as data vaulting providers who will sell data storage on their LPAR’d System i machines.
Chris Hird, Shield’s president, says the latest release of the operating system is very important to RAP/400. “The technology provided by i5/OS V5R4 is the key to the success of the product and how it can run automatically with no human intervention on the target system,” he writes in his blog. “Recovery options are limited in the initial version, but they will be improved as the technology grows and customers request such improvements.”
RAP/400 will gain more advanced functionality in subsequent releases, Hird says. “As we progress we hope to provide more object-level updates to the process, which will then bring it more in line with the current HA products but not compete in the HA space,” he writes on his blog. “There are already enough solutions out there for that.”
Hird adds that System i admins on a budget can learn how to build their own apply processes by attending a session at the upcoming COMMON conference in Anaheim, California, given by Larry Youngren, the IBMer who has championed remote journaling, titled “High Availability on a Shoestring.”
RAP/400 version 1.0 is available now. Licenses cost $2,000. For free 30-day trials and other information, visit Shield’s Web site at wwwshield.on.ca.
This article has been corrected. RAP/400 does not apply the receiver changes on a source link failure, as the story previously stated. It continuously applies changes every time the receiver is detached from the target journal. IT Jungle regrets the error.