Partner TechTip: Provide Continuous Data Protection for Data and Objects

High Availability / Disaster Recovery
Typography
  • Smaller Small Medium Big Bigger
  • Default Helvetica Segoe Georgia Times

Stuff, as they say, happens. An operator corrupts production data by accidentally loading the wrong dataset. An important file is inadvertently deleted. A query that was supposed to be read-only is mistakenly run in read/write mode, erroneously altering data. A developer modifies a program, unintentionally introducing a bug after the old, correct version of the software was mistakenly overwritten or otherwise destroyed. The list goes on, and those are just problems that happen by accident. When you add to that the possibility of malicious attacks that corrupt data and objects, the result is myriad threats to the organization's critical information and systems.

Tape backups are a common first line of defense against data loss due to disasters. And high availability (HA) solutions, which maintain hot-backup servers in remote locations, allow for continuous operations during planned maintenance, while also reducing recovery points and times almost to zero after disasters. But neither tape backups nor HA offer a practical way to solve the class of problems described in the opening paragraph. Recovering a chunk of data from a backup tape takes far too long because of the serial nature of tape and the need to retrieve it from its offsite location. What's more, as is described below, some intra-day data—possibly the data you'll need to solve a problem—doesn't exist on any backup tape.

HA doesn't offer a solution for this type of data and object corruption problem either because the HA replicator that maintains backup redundancy will copy any corrupted data to the backup server as soon as the error is applied to the production system. That is, after all, precisely what the software is supposed to do—ensure that the backup is a perfect replica of the primary server so it can immediately take over operations whenever necessary.

One option that does provide the means to correct the "stuff" that happens is a data vaulting solution with continuous data protection (CDP) functionality, such as iTera Vault from Vision Solutions.

CDP, also called continuous backup, is a generic term that refers to the automatic saving of all changes made to data and objects. In effect, because every change is captured, CDP keeps a copy of every version of all data and objects. This contrasts with tape backups that, nightly and/or weekly, save data as it was at that point in time. That's another reason backup tapes can't resolve individual data corruption problems: Because the granularity of saves is rarely finer than one day, you can't recover data to a point after the last save, nor can you re-create data as of some point that occurred in the middle of a day.

iTera Vault provides a solution. It uses the i5/OS or OS/400 journaling processes to capture data changes continually. These changes are then compressed and transported, at an interval defined by the administrator, to a separate backup storage platform that can be either local or remote to the primary system. iTera Vault also monitors changes to non-journaled objects and offloads each change immediately to the independent storage platform, appending them to the last backup containing the old version of the object. This is referred to as "snapshot-based" protection.

The backup storage platform does not have to be System i. It can also be a lower-cost, TCP/IP-accessible standalone PC, server, or SAN running under Linux, Windows, or UNIX.

Through a single System i–based interface, iTera Vault provides a high level of management for both the source System i and the alternate storage platform. iTera Vault's automated, rule-based intelligence manages data compression, transport, rotation cycles, capacity, and retention values to minimize operator intervention. Utilizing the timestamps that the journaling processes automatically apply to all changes, iTera Vault can also guide an operator through the process of recovering specific data to its status as of a particular point in time, thereby resolving the type of data problems portrayed at the beginning of this TechTip—problems that, given enough time, are unavoidable.

Because of the high degree of automation built into iTera Vault, it's possible to resolve data problems easily without the need for additional expertise, but, should an exceptionally complex issue arise, Vision's support group stands ready to walk an operator through the necessary recovery process.

See the MC Showcase Buyer's Guide for more information about iTera and Vision Solutions.

Joel Klebanoff is a consultant, a writer, and president of Klebanoff Associates, Inc., a Toronto, Canada-based marketing communications firm, and author of BYTE-ing Satire. Joel has 25 years experience working in IT, first as a programmer/analyst and then as a marketer. He holds a Bachelor of Science in computer science and an MBA, both from the University of Toronto. Contact Joel at This email address is being protected from spambots. You need JavaScript enabled to view it..

BLOG COMMENTS POWERED BY DISQUS