This article is part of an Essential Guide, our editor-selected collection of our best articles, videos and other content on this topic. Explore more in this guide:
2. - Enterprises, SMBs weigh the cloud DR choice: Read more in this section
- Are cloud DR services viable for the enterprise?
- When is it a good idea for SMBs to implement cloud DR?
- Weighing cloud DR against traditional disaster recovery strategies
- Enterprise IT execs question stability of cloud disaster recovery
- Is Disaster Recovery as a Service an easy sell with SMBs?
- For SMBs, cloud may be best choice for disaster recovery storage
Explore other sections in this guide:
- 1. - Why does IT need cloud disaster recovery?
- 3. - Executing and managing cloud disaster recovery strategies
This article can also be found in the Premium Editorial Download "CIO Decisions: CIOs discover effective business continuity and disaster recovery strategies."
Download it now to read this article plus other related content.
It's become fairly common for companies to use cloud storage to spin a temporary production environment up or down for testing and development or a short-term marketing campaign. On the other hand, enterprises are not exactly pinning their disaster recovery strategy on the cloud.
Many enterprises already have multiple data centers in place that can be used as primary data centers and backup disaster-recovery facilities. In addition, companies are concerned about security in the cloud, in terms of data protection and privacy, physical security, and application security.
Potential network downtime is another drawback when it comes to cloud disaster recovery. Enterprises can't put up with service interruptions regardless of whether the cause is a bandwidth constraint or a distributed denial-of-service attack. "It's all about quality, not about low-cost services anymore," said Lalitendu Panda, global CIO at D&M Holdings Inc., based in Japan. "Interruption of service is an issue; we have had a couple of 'situations,'" he said. "It's not like having your own [infrastructure] that you can modify. You have no control over what else is running on the cloud that could degrade performance."
Regardless of these concerns, the ease of using cloud storage is proving to be a draw for some enterprises. Cloud storage offers recovery assurances that are similar to those of other off-site storage options but cost less -- even though they provide faster recovery time objectives (RTOs), according to industry analysts.
"When this cloud concept came up, we were prepared," said Dan Zinn, CIO for the 15th Judicial Circuit of the Florida State Attorney's Office. His IT department had been deduplicating data to minimize the amount that needed to be backed up, and encrypting the data on tapes for a weekly rotation. With the cloud solution provided by CommVault Systems Inc. and Iron Mountain Inc., "scheduling a backup and clicking a button freed up the sysadmin's time, and gave me a solution so I didn't have to worry about [the tapes]," he said.
The heaviest users of cloud disaster recovery are small and midmarket businesses with annual revenue of $50 million to $1 billion, according to Dick Csaplar, senior research analyst for virtualization and storage at research firm Aberdeen Group Inc. in Boston. Enterprises with more than $1 billion in annual revenue typically have data centers in multiple locations, and have less need to use the cloud for disaster recovery, he said.
Last October, Aberdeen Group surveyed 100 organizations that have a formal disaster recovery strategy to learn whether they used cloud storage and if they did, whether they realized any benefits in disaster recovery performance. The survey found that organizations that had moved at least part of their storage to the cloud recovered four times faster than those with no formal cloud storage program. In addition, cloud storage users met their RTOs more often than those who kept data in-house.
The survey also found the following:
- Companies with a cloud storage strategy reported an average of 2.5 downtime events in the past year, and resolved those in an average of two hours. The longest downtime event among this group of companies, which had an average RTO of 12 hours, was 5.3 hours.
- Companies with no cloud storage strategy reported an average of 3.5 downtime events in the past year, and took an average of 8 hours to recover from them. The longest downtime event among this group of companies, which had an average RTO of 13 hours, was 13.7 hours.
A disaster recovery strategy was their No. 1 driver for using cloud services, respondents to the Aberdeen Group survey said. Those respondents deemed best in class, in terms of their cloud disaster recovery strategy, take the following measures:
- 55% deploy a secure connection to the cloud.
- 40% utilize server failover to the cloud.
- 22% do continuous data replication to the cloud.
- 10% use multiple cloud providers.
Tiered storage and IT governance
Cloud storage does not work well for data warehousing in a situation in which one database is accessed by several different applications, according to Andrew Reichman, senior analyst at Forrester Research Inc. in Cambridge, Mass.
It also presents governance issues that IT executives need to be on top of. "The problem is never with CommVault or Iron Mountain; it's always, 'Where is my data?'" said Zinn, of the Florida State Attorney's Office.
Interruption of service is an issue; we have had a couple of 'situations.'
Lalitendu Panda, global CIO, D&M Holdings Inc.
With the amount rising by the hour, data is tough to keep track of. At the Sisters of Mercy Health System based in St. Louis, clinical studies' storage demands are astounding. "Fifteen years ago, the first milestone was a terabyte of storage, total. Now we're doing that a week," said Jeff Bell, chief operating officer at Mercy Health, a network of 28 hospitals across a four-state region.
Gartner Inc. analysts project an 800% increase in the amount of data over the next five years. Of this data, 80% is unstructured and generally goes untouched after 90 days, according to Ray Paquet, managing vice president at the Stamford, Conn.-based firm.
"Storage is growing greatly," said Bell, who implemented a tiered storage strategy on an internal cloud. Older data gets placed on older disks, and data that repeats gets compressed. "Dedupe is big, so we don't have to back up everything," he said. The idea is to "store as few times as possible, put it on the correct tier of storage, and make sure it's available."
Done right, cloud storage promises to free up IT staff from complex and onerous storage management tasks. Nevertheless, CIOs should push for service-level agreements that are as good as, if not better than, ones they could offer internally, Forrester Research's Reichman said.
Let us know what you think about the story; email Laura Smith, Features Writer.