Spokane Regional Networking, Social Media, Professional and Business Development
Larry, it is true many companies I have talked to in and around the Spokane area do not have the proper set up for backups and have not gone as far as putting a Disaster Recovery plan in place. It is a good discussion to be having amongst business owners. What will businesses do if we have a disaster and their computers in their office are destroyed or not accessible?
Since "cloud computing" has many different definitions and it a hot topic right now, what form or definition of cloud computing strategy are you referring to? Are you referring to backups kept online via Amazon or someplace like Rackspace? Are you referring to virtualizing servers and data storage devices? Companies dealing with SOX, HIPPA and other regulatory systems have to be careful in the realm of "cloud computing" generally they have to take more traditional approaches to DR planning.
Are you referring to the following article: http://www.cio.com/article/677320/Embrace_the_Cloud_for_Disaster_Su...
Disaster Recovery does not look the same for every business. Questions to answer would be: What information and data do I need right now to get my business back in place? What information and data are not mission critical and can wait? Where are we going to work from? Where can we go to get business back up and making money again? What kind of business do we have and do we need to look at including a disaster recover plan that includes data being backed up to a location outside of the Northwest or are we allowed to keep our data with a place like Amazon? What resources and funds need to be allocated to getting the business back up and going in the event of a disaster?
Thank you for starting this discussion. It is a good discussion to have.
Great questions Katy. Many business cross their fingers and hope all is well and that they can resume business activites quickly after a disaster. The longer it takes an organization to recover, the less likely they will be around. There are some EASY, basic things any organization can do to insure timely recovery and resumption of normal activites. Our company specializes in many aspects of information management, including business continuity issues. We have been actively involved in this area for many years and have learned great lessons over time. We are happy to share our solutions and discuss keeping your business safe and sound for years to come, regardless.contact me or visit our web site for more. www.devriesinc.com
Business protection is like a seat belt in your car. Wear it all the time, because if you ever need it, you will be happy you did.
Another thing for the consultant or IT manager to be aware of the the need to understand the criticality of data they are backing up to the cloud. Generally cloud-based back-up is more reliable than tapes in off-site storage, in fact, people often discover only when they try to restore from tape after an event that their back-up was corrupt or incomplete. This is a good reason to go with cloud-based back-up. Be aware, however, that if you put 100Gb in the cloud, getting it back out is no simple matter; even with a decent Internet connection you could be looking at days, or weeks to download the back-up, slowing down regular traffic in the meantime. Some providers will offer to burn it to disk and mail it for an additional charge, but even this can take quite a bit of time and cause some serious inconvenience at least. Check out your local options for automated off-site back up; this will allow you to get to the data in person (or for the provider to run it over to you) if needed.
Having been in the Fire Service since 1979 I can tell you that Mother Nature has us "out gunned" so there is no such thing as "bullet proof".
So both high and low tech need to be considered. The biggest obstacle that I see is the money... and concessions are made. Mitigation has a much higher priority than prevention and recovery. If you can't survive the incident, recovery is a moot point. High tech recovery systems might sit there gathering dust and become obsolete before they are ever deployed. Low tech, although bulky, has the advantage of a longer shelf life and fewer infrastructure requirements.
Off site duplication in multiple formats would be ideal but we usually compromise for what is affordable and doable in terms of time and other resources. In terms of disaster preparedness and contingency planning, my experience is that we roll the dice to some extent. I wasn't in Spokane for the Mt. St. Helens eruption and subsequent fallout but our resources would be overwhelmed if it happened again today. Although we are in a better position to deal with it now than we were back then.
Any option that does not contribute directly to the bottom line of a business will have to be flexible, durable, and almost cost free. If that option is sharable from a mulititude of stakeholders that would help with the cost-to-benefit ratio. Is cloud technology there yet?
You raise a good question. The complement to it is Business Continuity Planning, of which DR is a sub-component.
All too many folks make 'assumptions' about what disaster recovery is. Consider the Flikr users who had their accounts deleted a few months ago. Many assumed that Flikr would restore their content along with their accounts. Nope. Flikr restored the service. But recovery, reloading all their pictures (and associated meta data) fell on the shoulders of the individuals. Nasty surprise.
Similarly, many folks think of disaster recovery as 'restoration of business.' Disaster recovery quite often is simply the restoration of a service, a business' physical storefront (for instance), or restoration of data after a hard drive fails. DR is very frequently a case of you get what you pay for, as well. Does your service provider ensure recovery in 2 hrs, 2 days, 2 weeks, 2 months, or is it even specified at all?
Business Continuity Plans go beyond DR. If your business suffers a calamity, does the business continuity plan switch you over to an alternate site, and have you up and running again (transparent to you customers) in 2 minutes, 2 hours, 2 days....you get the idea.
So the bigger question may be: How quickly can I resume business after suffering a business interruption of differing severity? Spokane's relative remoteness, relative to major urban areas, means these aren't trivial questions.
A few things we have learned over the years that still hold true even with the cloud offerings.
* Regardless of where you store your data, if you don't have a plan you are unlikely to recover from a significant data loss without a major hit to your business. Have a plan!
* Your recovery plan and off-site data storage need to be TESTED. We recommend at least once a year, preferably twice.
* While the cloud is often priced very well for storing data, you still need someone to call when things go wrong and nobody in "the cloud" is going to help you. Make friends with a local IT company and bring them up to speed on your plan, even if they aren't the ones storing your data.
* Learn about snapshots! Our system defaults to a minimum 17 snapshot, 6 month data cycle. Meaning we can pull any piece of data we have backed up for you in the last 6 months. Human error accounts for 1/3rd of all data loss. So if you delete something on Monday, realize it on Friday and your backups have overwritten 4 times since then...well, you're screwed.
* Customize your plan to your level of data reliance. If you can go 3 days working off paper receipts for customers and not lose any business your level of recovery is much less than an accounting firm in tax season that will lose thousands of dollars an hour if their server is down.
* Local backups (backups at your business) are almost pointless. Of the dozen or so ways you can lose data, a USB drive plugged into a machine covers about 2 of them. Redundant drives on-site is NOT data backup, it is hardware redundancy and does nothing for fire, theft, viruses, human error, power surge etc etc.
* Get a system that tells you when backups DON'T run, not when they are working properly. Time and time again we see systems where everyday it sends an email that says "your data backed up"...and those emails start getting deleted without being read after about a week. If your system tells you when its NOT working, you will pay attention.
Here is a sheet I've been working on that talks about some "levels" of backup. These are arbitrary but valid in their differences. www.ilinkadv.com/backup.pdf
Good response items. I can say, I've experienced the issue of backups that became corrupted, only to not realize it until too late.
As a SaaS product manager in the past, I can definitely say being able to -test- fail-over systems is important.
Your comment about alerts is also appropriate. Alerts that are triggered too easily, too often, are like the boy crying Wolf! They quickly become so much spam, readily ignored, and achieve nothing. Guidance I've given teams: measure what we need to, to do our job. Don't measure 'everything' just because we can. That is is very true in this context as well.