This Blog has been moved.... go to www.verologix.com

Hello There,

This blog has been moved to our website, www.verologix.com.  Its called Rethink IT.

Enjoy....

--- Marc Potter

No Disaster Recovery, No Excuses… New Strategies that make it affordable

Do you keep your life savings under your mattress? Of course not! I’ll bet you don’t keep all your company cash in your desk drawer. Then why would you put your business at risk by not having a disaster recovery plan for your data?


What is Disaster Recovery (DR) in a nut shell?

Disaster Recovery is the ability for an organization to recover business systems and restore data to support business operations following a disruptive event. A disruption event can be as simple as tripping over your server cord and cutting power, so don’t think it can’t happen. Just refer to disruptive events of Hurricane Katrina, 9/11, and even the completely bizarre weather on the east coast this winter 2010 as very real scenarios where DR plans were deployed in a, “this is not a drill” fashion. Still don’t believe me? According to research by the University of Texas, only 6 percent of companies suffering from a catastrophic data loss survive, while 43 percent never reopen and 51 percent close within two years.

Here is my disclaimer!

Disaster Recovery is a huge topic that has many subsets of Risk Management and Business Continuance. I am strictly focusing on DR as it applies to IT, data protection and recovery. If you are tasked with putting together a full blown DR plan for your organization, this is a good start, but do not overlook the other critical components of enterprise risk management, physical DR and employee displacement, temporary facility plans, etc. A full DR plan extends beyond IT data protection and includes your entire organization.

Disaster Recovery (DR) is often mistakenly confused with High Availability (HA), which is a term used to define the amount of acceptable downtime a company can accept. Simply put, DR is, “can I recover my essential data and business systems”, and HA is “How fast can I recover my critical data and business systems”.

The old Paradigm broken

The simple idea of DR is nothing new to the large enterprise. The problem is, under the old paradigm, DR costs can be viewed in terms of millions of dollars. In the ancient days of IT dating back to the early second millennium, it was very costly to put together a DR environment. An organization would have to replicate the entire infrastructure at the DR location from servers to switches, applications, firewalls, appliances etc. This left many companies with very generic Dr methods, some only as advanced than a tape backup sent offsite and stored in the IT manager’s underwear drawer.

The new Paradigm… (Strategies that make it affordable)

Here are a few examples from the “very simple” to “very reliable yet affordable”.

Key: SOHO (Small Office Home Office), Small Business (Small office with fee employees), SMB (Small medium business with several to many employees), SME (Small medium Enterprise with many employees over large geography), Enterprise (Very large company, global)

Simple Backups for DR (Minimal DR)

Recommended adopter: SOHO, Small Business

Cost: No Excuses cheap. If you can’t afford this, shut your doors and find a job, you’ve probably spent more on your cell phone.

Overview: An External HD’s and your operating system’s built in backup utility will give even your kid’s lemonade business the ability to have simple DR.

The Online Backup

Recommended adopter: SOHO, SMB, some strategies for SME, Enterprise branch offices

Cost: Very Cheap to Very affordable

Overview: Online backup services are an affordable way to add DR to your business AND automate your backup process. Smaller organizations can virtually “set it and forget it”. Gone are the days of constant babysitting of manual tape backups. For larger companies, online backup strategies can be deployed to backup hard to manage endpoint systems such as remote, branch office or laptop fleets.

Credible online backups services will offer secure, encrypted, semi-hands-off backup services over the internet, (aka in the cloud), to a co-location facility. These “business-class” backup services will also have fully secure and compliant, redundant systems that use high end enterprise class DR and HA infrastructure and methodologies that most smaller enterprises could never afford to deploy.

Note: It is highly recommended to avoid consumer grade online backup services for your business. They do not have the security, availability and enterprise features mentioned above to properly protect your data. Losing your mp3’s is one thing, losing your intellectual property, client lists, or accounting data is another entirely.

Enterprise DR and replication with virtualization and other cool new tools.

Recommended Adopter: SME, Enterprise

Cost: Considerably less than DR solutions of the past, and with the added ability to add HA for far less cost than mirroring your data center to two locations. A simple replication system using virtualization can start from as little as zero dollars, (using free tools like Citrix XenServer and two existing servers). Think of it this way. At the time your lost data becomes valuable enough to put you out of business, these DR strategies becomes delightfully affordable.

Overview: Recently, methodologies and concepts around virtualization, cheap and smart storage, and cloud computing, have created a paradigm shift. These concepts have given us the ability to remove data and workloads from hardware, and easily move them around to agnostic server configurations across geography, creating affordable and downright cheap Disaster Recovery.

Some vendors to look at include VMware, Marathon Technologies, Novell’s Platespin, Double Take, and Stratus Technologies. Smart Storage and SAN vendors to research may include Compellent, Dell Equallogic, Falconstor, Starwind Software, Data Core, and Dot Hill.

Next Steps

There is no such thing as a “one size fits all solution”, so be cautious of anyone who tries to sell you on a solution without a proper consultation and discovery of your business objectives. There are many different solutions and strategies out there to choose from, so take your list of excuses and roll them into a ball, wrap tape around them and use them for office baseball. That’s all excuses are good for anyway, right? Remember, there are no more excuses. Protect yourself. Protect your data.

5 Virtualization Strategies for the Budget Challenged and Reasons to Start NOW!

Every so often a technology gains enough inertia and acceptance it changes the game for everyone.  When this happens, there is a certain buzz in the air for guys like me. It’s because we know that the top players will be burning the midnight oil wrenching on new great features and products to fill the space.  The pricing gets more competitive, and the industry opens up to a whole bunch of “trailer hitch” companies that bring some much needed value to the table.  Companies like Platespin (recently purchased by Novell, Marathon Technologies, Vizioncore, and others all have great offerings that extend the virtualization story.  Yes, Virtualization is that game changer that has geeks giving up star trek at 2 in the morning.

With the great buzz around the technology, why are some companies not on the “V train”? Great question and one I keep asking them.  In some cases, it is education or lack of good information, in other cases, it is budget related.  When departments are all fighting over the same “budget bone”, how do you get your decision makers to Act? Give them some great virtualization strategies that will show them how to pay for it.

A little background

I’ve heard some talk that this technology is too new to trust mission critical applications with. Truth is, the premise has been around for over 40 years and first used in mainframes in the 60’s.  Virtualization as we know it on the x 86 platforms was championed by VMware in the 90’s.  Today, enterprise virtualization has very little overhead, and virtual servers run with little to no performance impact.  The reasons for using virtualization are well documented.  It can reduce CAPEX (capital expenditure) by eliminating the need for physical servers by 10:1 or more. The byproduct is a large reduction in electric, cooling, and physical footprint that can reduce OPEX (operations expenditures) by reducing downtime, and increasing productivity.  Additionally, Virtualization can help reduce risk, and increase uptime with live migration, workload balancing and other features that can increase application performance and ultimately end user experience.

If I can help just one starving company by introducing low to no cost virtualization strategies, then my job is done! Not really, but I always want to say that.  In reality, there are no good excuses for not at least putting together a virtualization strategy.  The following are 5 virtualization strategies for the budget challenged company.

Strategy # 1: The Starter Strategy.

If your company would like to get started with virtualization, but you think you cannot afford “enterprise class” virtualization, you can!  Thank the highly competitive market for virtualization hypervisors for that.  Know that there are some “trade offs” to consider when choosing your free hypervisor.  VMware ESXi a great choice for shops that plan on using VMware management tools in the future.  Citrix XenServer 5.5 is a very complete offering that includes centralized multi server management, and live motion among the features.  Microsoft shops using Hyper V can use the free Citrix Essentials for Hyper V Express edition, which adds some needed enterprise functionality. Novell Suse Linux Enterprise and Open Enterprise Server have built in virtualization based on the XEN hypervisor with some Novell mods optimized for MS Server, Linux and NetWare guest OS’s. (Note the Interop agreement between Novell and Microsoft ensures high performance between Microsoft and Linux). Don’t be too gun shy when choosing a free hypervisor, you can easily change hypervisor platforms using V2V tools to migrate virtual to virtual platforms. Platespin is arguably the best, but VMware, Citrix and other 3^rd parties tools exist, at no cost. 

Strategy # 2: The Heterogeneous Hybrid Strategy

Early adopters of virtualization in many cases are using VMware as their vendor of choice.  They are the 800 pound gorilla that owns the market share, and the premium licensing costs that go with it. In all fairness, they deserve the position, but in some cases, there is an argument for a heterogeneous strategy.  The hypervisor has pretty much been commoditized.  The battle for supremacy really boils down to virtual machine management. If you study a comparison guide of virtualization platforms side by side, you will find that VMware, Citrix, Microsoft, Virtual Iron (now Oracle), Novell, SUN all can consolidate servers, isolate workloads, and can be managed in some way between physical hardware.  Did you just buy the top of the line platform for a 30:1 consolidation when you only have 10 workloads? What did you pay for that? Do you have a mission critical workload that you really want VMware ESX server, but you can’t afford to do it this year? Move some non critical workloads to a free hypervisor, like VMware ESXi so you can still use the VMware management tools.  Do you have a Citrix XenApp server farm? Move those to Citrix XenServer 5.5 which happens to be optimized for Citrix XenApp Servers. 

Let me give little warning here.  Companies who are planning a whole hearted leap into VMware vSphere should know that this is a very complete, integrated strategy that has some state of the art technology, and is NOT very friendly to the heterogeneous strategy. It can be done in doses, but know you may be losing some vSphere functionality.  This is a trade off that must be decided by each company.  vSphere is not a budget challenged discussion, it is a performance and optimization discussion.

Strategy # 3: The Extending your End of Life Systems Strategy

How many times do you hear the phrase, “If it ain’t broke, don’t fix it?”  Well this strategy takes that to heart.  Do you have an old proprietary program that does everything you want, because you built it in house, you love, it works it runs your multi-million dollar manufacturing environment…. But it is on a dying platform on dying hardware?  This scenario brings a tear of joy to the eye of the sales person who gets to convert you over, but this strategy can help keep ‘ole data base betsy working hard for a little while longer and putting the conversion timeline in your control.  

The main goal is to at least get you working on up to date hardware that isn’t ready to blow up at any moment. Before virtualization, your hardware would dictate which operating system you would need to be on, and by default, which data base would be supported, on down the line to “obsoletsville”.  You’ll probably find allot of these systems were built in the 90’s to run on NetWare, a very bulletproof platform that just works.  (VMware ESX server, and Novell OES 2 server both can support NetWare guest OS’s, check each companies compatibility list for specifics.) A point of note, Novell is still a very strong company with some very good products.  This claim is strongly supported by the Microsoft / Novell interop. See www.moreinterop.com so don’t be talked into a premature rip and replace when a little strategy will go a long way.

Strategy #4: The Poor Man’s Disaster Recovery

Disaster Recovery (DR) is a very big subject. The simplest being, backups, the most complex including physical recovery of assets, people, over geographies, blah blah… too much to cover here, BUT what virtualization CAN do on a shoestring budget is simple replication of workloads to a safe place. 

A workload is the VM, OS, and Application.  Consider this strategy mission critical if you don’t have DR in place. It gives you the minimum protection that really can’t be done with simple tape backups. Simple replication of VM’s using live migrations tools will also get you prepped for stepping up to better solutions like Platespin Protect or the Forge hardware appliance, or VMware’s Site Recovery Manager or Marathon’s everRUN which automate the DR process and provides better protection and availability.  Before virtualization, DR was a 6 figure and up buy in with some enterprises spending millions of dollars.  Today, it can start for free, and scale up to budget worthy.  The best free option here is Citrix XenServer 5.5 since it comes with live motion and multi server management, and later when paired with Marathon eveRUN (a strategic Citrix partner), you have a top shelf solution that is a steal compared to the alternatives.

Strategy #5: The Storage Virtualization Strategy

I know this one may be considered a little out of scope of the virtualization conversation, but storage virtualization is being leveraged by many major vendors to provide some real measurable cost savings.  Multimedia applications are quickly becoming part of the normal enterprise workplace, and videos, pictures, and .mp3 content is eating up disk space on email and web servers.  

Like server virtualization, storage virtualization represents physical storage devices as software, enabling them to do all kinds of wonderful things like central pooling of resources, thin provisioning, replication and more.  Storage companies that are leading the field today such as Compellent, Equallogic, SUN, HP’s Lefthand SAN all use storage virtualization in their products in some way to achieve their functionality.  Many SAN solutions are not intelligent, but JBOD (not my definition) is just a bunch of disks.  It should be “Just a Bad Optimization iDea”.  They end up costing a lot of CAPEX, and OPEX dollars. 

How does this help my budget you might ask? Well first, the cost of intelligent storage has come down recently. You can now start with a smaller iSCSi SAN at a lower starting point, add central storage management and thin provisioning very inexpensively. Software based companies like DataCore, StarWind offer software SAN solutions that can be built on your own hardware and JBOD disk arrays for a song.

Wrap it all up with a bow

Virtualization should be something affordable to all companies.  Let me go out on a limb here and make a statement. Except for very specific instances, all servers should be virtualized. Some companies today are starting to adopt single instance virtualization to take advantage of the administration and management benefits.  Small companies that would like to add a new application, but not the cost of a new server can now do that, and even mix the Operating Systems.  Imagine the newly found freedom to now add a open source Linux program to your once “under utilized” windows server.

With strategies that are as budget friendly as FREE, there are no more excuses.  Virtualization new comers… Find a non mission critical workload and virtualize it with a free solution.  Build a template and see how fast server provisioning can be. Use P2V tools to see how easy it can be to move a physical machine to a virtual one. Then put a good virtualization plan together, educate yourself, find help where needed, and reap the rewards… within your budget.

3 Reasons Identity Management is a “MUST”… and strategies to make it affordable

Let me be clear about what I’m saying. Identity Management for companies with employees that have access to critical data is a MUST, not a want.  Over the years I have had the privilege to work with many companies large and small, who have different business needs.  In many cases I hear all the reasons, (and sometimes excuses) for not implementing a solution, policy or methodology. Sometimes these reasons even make perfect sense!  In making any business decision, the choice to do, or not do anything is weighed by what I call the “risk vs. reward scale”.  Regarding Identity Management (IDM), if you have employees with access to critical business information, you MUST put at least basic IDM in place!

So what is Identity Management? Bill Brant, CEO of Directory Service, Inc. says “IDM is the technological automation and enforcement of business policies and processes to manage the lifecycle of electronic credentials, entitlements authorization and compliance mandates.”  If you are in management like me, let me translate in English. IDM automates your logins so your company is secure, and you don’t lose millions of dollars, PLUS it increases productivity so you can make millions of dollars. The following are my top three reasons IDM is a must, not a want.

Reason One (1):   Provision of new employee credentials

Companies that do not have Identity Management spend days to weeks to properly provision a new employee, and with a high probability of improper provisions.  The popular method used to accomplish this task is a simple email request.

Typical email thread:

HR to IT Admin:  “Jack is starting today with us, can you get him a login?”

IT Admin to HR: Sure what does he need?

HR to IT Admin: “He is working in Sales, ask his supervisor.”

IT Admin to Supervisor: “Jack is starting today, and I need to get him a login, what accesses does he need?”

Supervisor to IT Admin: “I don’t know, how about just copy the access rights from Jill, she’s been here a while, so whatever she has must be right?

Risk to the company:  Jill was the Engineering Manager and Marketing Supervisor before becoming the top sales person in the company.  Each new position gave her role specific rights that were never properly taken away as she changed roles.  Now she is being used as the “template” for user rights to new hires.  Jack the new hire, just gained access to engineering blueprints, and new “go to market” strategies. In addition, the back and forth emailing took two weeks because the supervisor was on vacation.  Adding a face slap to a poke in the eye, Jack the “new hire” is still being paid even though he had no access to do his job. Sound familiar?

IDM to the rescue:  A company with IDM could implement automated provisioning of credentials by role.  A company would define the accesses any given role can have, and further, lock out accesses for roles they should not have i.e. the janitor does not need access to the accounting system.  The IDM system’s automatic provisioning process tool performed this task in seconds, and Jack was properly provisioned before he sat at his new desk.

Reason Two (2): Deprovisioning of terminated employee credentials.

In a company without Identity Management the same situation occurs as in the scenario above, but with more immediate consequences.  The popular method of conducting deprovisioning of credentials in a company without Identity management is by way of a simple email request.

Typical email thread:

Supervisor to HR: “Jack has been terminated immediately for bad attendance. Please put all the termination protocols in place. He has been removed from the facility, but he did not have his badge with him.”

HR to Supervisor: “Out of Office Reply” I’m sorry, but I’m out of the office the next two weeks on my honeymoon.  I my absence please contact the supervisor”.

Supervisor to Manager: “I just fired Jack, and need the termination protocols, but HR is out of the office, what now”?

Manager to Supervisor:  “Who is her Backup in HR?”

Supervisor to Manager:  “I am, but I don’t know the protocol.”

Manager to HR:  “when you get back from your honeymoon, please terminate the supervisor, he hired Jack who we think may have stole engineering plans and sold our marketing plan to the competition after he was terminated because he still had his accesses for the last two weeks! Of course we cannot prove it.” (side note to reader Yes IDM applies here too for compliance and auditing, but that is another article… Marc).

Manager to CEO:  “I have no idea how our engineering blueprints and our marketing plan got into the hands of our competition?”   It must have been Jill, she has rights to both of those areas. By the way, I’m hearing our client list is being aggressively called by our competition as well.  It couldn’t have been Jack, he’s been fired for weeks now.”

Ok obviously I was on a little bit of a roll there with the Manager reply, but I think you get the picture.

IDM to the rescue: A company with IDM could implement automated deprovisioning of credentials by Identity.  In this scenario, Jack could have been deprovisioned before he was even out the door.  If he tried to access his client database from home, he would have been locked out.

Reason Three (3): Identity Synchronization and Password management

Did you ever think that 3M would produce the world’s largest and most used Identity Management and password vault tool! It is true! Its call the “Post-IT” note, and it can cost you millions.

Some people may get basic Directory Services and Identity Management confused. Directory Services are a key part of IDM because this is where the Identities are managed. For example, Active Directory, eDirectory, LDAP, are all network directory services.  What about your applications that maintain their own “directory service database? This may be your custom built Inventory application, or ERP system for example.  How do you get these systems to talk?  If you do not have Identity management, you create separate login credentials for each sub system, and have your end users become the (Identity Management).  This becomes the Identity Management by “Post-IT” note that was mentioned earlier.

IDM to the rescue:  With IDM, companies can synchronize their user passwords between directories and application directory databases giving your end users a single password to manage for all systems.  The next step would be to implement SSO, or single sign on, which automatically uses a single login event to sign into multiple databases eliminating the need to manually login to multiple systems many times. I stop short of saying SSO is a “MUST” for all businesses, but it sure is up on the list of “should haves”.  I reserve the right to be on the fence on the “SSO vs. Identity sync only” discussion depending on the client needs.

Password management is bundled into this category, but I could add this to the list on its own.  Some may argue that this is not IDM because it is a directory service component, but I believe it is a component of IDM, so take it for what it is worth.  Password management in this scenario would be more than just enforcing strong password policies; it would include “self service password” assistance using challenge response questions and secure authentication methods like multi factor authentication and one time passwords.

Strategies that make it affordable:

There are many different products out there that can facilitate Identity Management and Access Controls. Some of the best are made by Novell, Sun, Oracle and IBM.  Recently, the Identity Management space has become somewhat commoditized in what I would call the “basic IDM” space. This would be the space I touched on today, with provisioning / deprovisioning, password management, and synchronization of identities. Some of this functionality is being built into the OS and Directory Services of some vendor products from Novell and Microsoft.  Novell has Domain Services for Windows, eDirectory, and the IDM bundle edition that ships with Novell Open Enterprise Server 2 (OES2).  Most major directory services vendors have free self service password management tools available for eDirectory, Active Directory, Sun Directory Server etc).  New companies are building targeted IDM solutions based on open source like, GreyTower from Directory Services, Inc., and Sun. These solutions can be implemented without licensing costs, but also sell support and maintenance if you need it.

Take the first steps. Contact your trusted Identity Management advisor and discuss your options. Make sure they are not tied to any single vendor or you will get a single option presented that may not fit your business.  Remember, IDM is a MUST!

Backups 101

Backups 101

Backups have one purpose. Data protection.  That said there are several considerations for data protection. 

- What is your data worth to you?

- how long can you be in business without your data?

- How much does it cost per hour / day to be without your data?

- in what format do you require your data to be in? Paper vs digital (yes paper is data)

- what is your retainment policy IE how long do I need to keep my data aka archives. (another topic: Data Archives including email archives have become a major pain point for companies because of legal ramifications and requirements for eDiscovery and compliance should an employee sue your company)

Because most people need data, in digital format readily available, we have backup systems.

Brief History

Backups are really just an extension of data protection methods put into place to protect data on paper.  In the old days, data was created by typewriters or printed on massive green bar reports by the "data processing" department.  Even though the data was warehoused on massive reels of tape, it was typical for this data to be "backup up" by printing it all off on paper and send off site for disaster recovery. 

The last 15 to 20 years, the typical method of backing up data has been using automatic or semi automatic tape backup systems.  Depending on the budget of the company, there would be either a single tape that required manual intervention or a tape library that could automatically change tapes as required.  Early tape backups were analog reel to reel, then later digatal tapes like DAT, LTO formats came along that gave better performance and larger capacity. The main reason for using tape was the cost. Hard Drives were very expensive per Mega Byte.  Recently, with the cost per mega byte going way down, newer technologies that leverage D2D or "disk to disk" backups have become the best method of backup.  A D2D2T or "Disk to Disk to Tape" method can also be used to simplify and automate backups, reduce backup times, but still allow for tapes to be taken off site for DR (see below).  As the Internet has matured and bandwidth has become cheap, online backup systems and methods of replication off site have become the popular. This takes D2D to the next level by eliminating the need for antiquated tape systems altogether.

Disaster Recovery aka DR

Data Backup is only one piece of data protection.  Disaster Recovery is the next.  In short, "disaster recovery" as it relates to data backup is the method of ensuring your data is available in the event of a disaster, regardless of the size of the disaster. This can include a fire sprinkler flood, theft, coffee mishap etc.  In most cases this is simply the act of taking your data off site.  In the old days this was one person who had the job of manually ejecting and schlepping the tapes home.  Today, it is leveraging the Internet to "automagically" copy or replicate the information to a safe place off site.  The added advantage of this is the protection from regional disasters, for example an earthquake or hurricane where the entire region may be at risk, and yes, the tapes you took off site to your house are still at risk.

A sub topic of DR is High Availability HA  or business continuance. This means, how fast can your systems get up and running in the event of an outage or disaster.  Most basic backup systems do not provide for High Availability (this is also a new topic regarding server clustering andvirtualization strategies).

Summary of recommendations

For the typical small business or SOHO

Check for online backup vendors that can give automated backups securely over the Internet. These methods backup to hardened facilities with Enterprise level data protection on the back end.  Remember to test these systems periodically.

Smaller companies can consider services like Mozy, or carbanite, or of course, you can just ask your Verologix rep to point you in the right direction.

Small to mid and some Enterprise companies with more mission critical data should consider a system that backs up locally first then offsite, such as Barracuda Backup appliance and online service.

If you cannot live without your systems, contact your verologix rep at www.verologx.com or call 800-403-8041 and ask about DR with High availability and virtualziation options, but be prepared to spend a few bucks.  these designs can range from several thousand dollars to over ahundred thousand dollars for extreme DR and HA.

If your brain is bleeding, feel free to ask for help!  

Blog in Transition...

Hello,

I decided to separate busienss from non business blog stuff.  On occassion, I blog about anything that I think is worth it at wesstess.blogger.com

After posting a few things I thought were great for my industry, and then thinking... these tech people don't give a rats tail what I think about global worming... maybe I should separate these!!!

... At the end of the day, does it really matter? I think I may be the only one writing and reading these anyway...  hmmmm :-|

Please let me know your thoughts... 

-- mrp