6 New Tools Every SEO Should Check Out

Posted: December 16, 2010 in SEM, SEO, SMO

Great post from SEOmoz…

*****

 

There’s some terrificly useful new tools on the market that very few SEOs are aware of or using (at least, if my experience is any indication). It’s my duty, therefore, to share with some of these shiny new sites and let you explore, engage and apply to your campaigns and efforts. Hopefully, these will add great value for you, and expose them to folks who really need their help.

Ontolo: An Expert Link Builder’s Dream

Link building is hard – really, really hard. Ontolo tries to productize many of the manual tasks, searches and tracking processes of link building with an extensive, diverse toolset. You can see a big list of their link building tools here, everything from .gov/.edu finders, to competitor link searches, to content research and backlink tracking/monitoring. They even have some nifty “productivity hacks” (small, simple tools to help with menial tasks like combining keywords or removing duplicates from lists).

Below is a screenshot of Ontolo’s Authority Links searching tool. As you can see, there’s a multitude of options including clever sorting/filtering systems.

Ontolo's Link Finder Beta

Ontolo isn’t for everyone – you should be a professional or semi-pro link builder (the tools and their results, rely on a lots of prior knowledge), but for those it’s geared toward, the reviews have been outstanding and Ontolo’s team (Ben Wills + Garret French) are constantly upgrading the service and functionality.

_

SEO Gadget’s Keyword Research Beta

Richard Baxter is constantly lauded as one of the SEO industry’s best, brightest and most driven minds. I curse myself for not smuggling him into the US, forging documents so he can stay, slapping an American accent on him and chaining him to a desk at SEOmoz (OK, maybe that’s a little extreme).

Luckily for you, my evil plans are for naught, and Richard’s talent has born fruit for all of us in the field with his remarkable new keyword research tool (currently in beta).

SEO Gadget's Keyword Tool

You create campaigns (similar to the SEOmoz web app), plug in your Google Analytics account, and sort keywords into relevant groups. The tool then lets you visualize potential opportunity for keywords you’re already ranking for and those you haven’t yet targeted by combining rank data, traffic data + search volume data (via Google’s AdWords API). It’s a brilliantly useful tool for those seeking new ways to ID the keywords that matter and take action on the low hanging fruit.

_

Trunk.ly: Terrific Tracking for Your Twitter Links

Trunk.ly is a deceptively simple, currently free, application from the brilliant minds at BinaryPlex (who also built the much more full-featured Tribalytic, an Australian-focused social platform for measuring influence and share of voice).

Plug Trunk.ly into your Twitter account and you’ll get a page like the one below that shows a timeline of the links you (or your friends) have tweeted.

How is that useful you say?

Because people tweet some pretty brilliant stuff and they show, through tweets, what they care about. Whether you’re relationship-building with a new contact, seeking topics for linkbait or content creation, attempting to determine the impact a particular individual has on clicks/rankings via their account or just interested in what someone has to say, Trunk.ly’s a great way to do it.

I’ve been surprised at how often I use it just to find the things I tweeted!

(BTW – Trunk.ly’s currently in private beta, but AlexTim, the creators, may have some invites if you leave comments in the post or tweet at them)

_

Markup.io: Flawlessly Simple Screenshots + Notes

You know what’s a pain in the butt? Taking a screenshot of a page, pasting it into Photoshop, adding notes to it, saving the file to be small and emailing it as an attachment to a developer/designer/marketer/manager/co-worker.

Markup.io to the rescue!

Markup.io on the NYTimes Most Read

In a stroke of simplicity made brilliant, markup.io lets you drag a bookmark to your broswer that will, on click, show an overlay that lets you “mark up” any page on the web with text, lines, arrows, boxes and circles of various thickness, size and color. From there, you can directly export or share the resulting screen image. It works so intuitively that our team at SEOmoz has been finding massive utility and time savings by employing it vs. a manual process.

_

Content Optimizer: Leveraging LDA for Keyword Suggestions

I already knew Russ Jones and the team at Virante were pretty smart, but this time they’re just making me look bad :-)

One of the biggest frustrations with our Labs LDA tool (warning, it’s still in super beta and may not be particularly performant if lots of folks are actively using it) is that it doesn’t recommend words and phrases to add or take away from a piece of content to help make it more “topically relevant” to a query. Building that would require a ton of very hard computer science work… Or would it?

Content Optimizer from Virante

The team from Virante got impatient (a great trait in any startup) and built their own recommendation system. It’s not perfect, but in many cases, like the example above, it can help. Basically, it grabs the top results from Google, checks for words those pages have that your page doesn’t and runs LDA scores for your page (through the labs tool) with and without those keywords. Pretty sweet hack, right?

_

Andrew Wright’s + Ben Huff’s SERPs Analysis Worksheets (MS Excel)

A couple months ago, I presented a methodology of how to understand the ranking algorithm for an individual search result graphically to help figure out why the top ranking results are ahead of those below them (and what metrics you might need to tweak/improve to reach the top). It’s certainly not a perfect strike, but many folks were excited and interested.

In fact, Andrew Wright, one of the crackerjack consultants from Bloom Media in the UK, put some serious time into improving my models and the Excel spreadsheet to make it more usable, understandable and useful.

SERPs Analysis from Ben Huff

SERPs Analysis via Andrew Wright

But, that’s not all. SEOmoz’s own Ben Huff looked over Andrew’s work and had some tweaks of his own. In the Box.net folder I’ve linked to, you’ll find both versions of the Excel worksheet. By Q2 of next year, we hope to have this SERPs analysis system included in our web app (so they’ll be no need to go crazy in Excel). In the meantime, though, we invite you to check out the work of these two, building one on the other (Ben Huff worked on the initial version with me for the face-off in London) and give feedback. It may not be the most scalable way to analyze a search result, but currently, it’s the most comprehensive and likely to produce a good answer.

 

The following post is from Alec Cochrane’s blog.  Alex used to be the Web Analytics Manager for a large B2B publishing firm. Now he is the Web Analytics Manager for a Government portal. Alex writes about Web Analytics stuff and related subjects (Usability, SEO, IA, etc).

*****

I’ve just realised it has been a month since I last posted.  That is too long, so my humble apologies.  For those of you who aren’t aware I have been guest posting on eConsultancyas well (follow me on facebook/twitter if you want to get all the updates!), so I have been busy still.  Anyway, moving onwards, I thought today I’d look at a little side project that I’ve been working on looking at a website where the conversion takes place off site.  This adds a level of complication that you spend half your life looking at solving.  This is especially pertinent for those sites whose sales process is completely offline (eg lead generation websites) or where the sale process is online but on another site (eg PayPal, eBay, Amazon, etc).  Here are my top five tips on how to get better analytics:

1: Collect as much data as you can

This may sound blindingly obvious, but if your users complete the sale offsite, collect the last possible point that they’re on your site as your conversion point.  By that, I don’t mean the last page (because many pages have a ‘checkout’ link that will point out of your site).  What you can do instead is tag up the outbound link so that you can collect the data.

In Google Analytics you need to collect the information as a faux page view using the_trackPageview parameter.  This effectively reloads your tag when a user clicks on a link with a custom value for your page url.  I’d recommend you choose wisely as you don’t want to use a link that is later going to be used as a proper page.  Google has a very good example of how to do this on their support site:

href=”http://www.example.com” onClick=”javascript: pageTracker._trackPageview(“/G1/example.com”);”

This onclick event is something that can be used in SiteCatalyst as well to much greater effect.  As well as being able to set up custom links on any page, you can also do the same in your downloads reports and your exit link report.  You can also include in your onclick event LinkTrackVars and LinkTrackEvents (look them up in the help under codes 1452 and 1453 on how to complete these from a technical point of view).

I’m sure that all other tools have something equivalent you can use with an onclick event – talk to your account managers and they’ll be able to tell you.

As with all of these work arounds you need to be wary that you aren’t measuring your end goal.  This is just the last available point you can collect data to.  What you would see if you could look at the whole data set may be that different traffic sources, partnerships, etc convert in different manners – but you have less control over that.  What you do have control over is your website and how many of them you can get across that divide.

2: Use all the data that is collected

Frequently I find people who have collected a whole host of data on one system not comparing it to a whole host of data they’ve been given from another system.  The most frequently used response to this is that they don’t match up and you are comparing apples with oranges.  Those of you who have been paying attention may have noticed that I even said as much in my last post - you can’t compare analytics systems.

Just because you can’t compare analytics systems doesn’t mean you shouldn’t use both of them.  They can both be used for various different insightful bits of information that can then be used together.  The trick is to compare the insights and not the data.
Early Analytics experts showing that there is a link between conversion and rotation of the screen
That is to say that if you have a lot data about your sales process from the point that they leave the site you can easily compare this with the information that you have about what people do on the site.  A promotion driving more people to click through to sale, but sale conversion has gone down in the same time then you may be able to link the two and do something about it.  Think about the things that you’d like to do about it and see if they make an improvement.  This stuff is all about insight – use as much as you have available in your arsenal.
When looking at lead generation, this should be much, much easier.  You should be able to track using other systems what goes on between collecting the lead and making the sale.  It may take several months of course and you may need to get your company at a stage where they are monitoring things like how many phone calls they are making and how many sales based on medium (web, phone, face to face, etc), but it should be possible.

3: Collect more information in your web analytics tool

For lead generation, you tend not to be able to collect information in your analytics tools about sales, but where the sale process is online, it may be possible.  In fact, it is the way that affiliates and aggregators have been paid for years.  The user does all the fancy stuff on one website and when they buy off the second website a cleverly inserted tag on the final page lets you know that they have bought.  Everyone is a winner.  Well you are.
Suddenly you have another step in the process because you can match like with like information on where people have come from, what they’ve done, etc.  More importantly the website that you are selling through should be more than delighted to do this because it means that you’re more likely to improve your site to be able to sell through them (thus making them more money).  You may find certain difficulty in this though where big websites like Amazon and eBay insist that you use their information, but if you can negotiate with them then it may be more likely.  Insurance websites, in my experience, are much happier to deal with super affiliates that affiliate networks because they think that it gets a bit closer to the customer.
Even doing this sort of thing as an affiliate has its possibilities.  If you can auto generate a random code from your analytics tool when people click pass the details through (a ‘Customer code’) that you can pass into your third party tools then you should be able to link up the data sources directly.  They won’t match because of the different filtering you do, but you a mapping exercise will allow you to look at like for like.

4: Collect more information in your sales funnel tool

So you can’t collect the information in your analytics tool.  Can you collect the information in your sales funnel tool?
This is probably more pertinent to lead generation websites, but why not pass into your sales funnel tool some information about how the user got to the site in the first place?  It should be a simple process using cookies to store a users campaign code if they arrive at the site and then pass that through into your sales funnel tool.  That’s what analytics tools do anyway so inserting this simple step shouldn’t be difficult.
This allows you to do something slightly different to the above – it takes away a huge amount of information about what the user did, where they came from etc, but for your campaigns you should get a whole host of information about how likely they are to convert.  This will allow you to prioritise your spend on different areas.  It won’t give you everything that you wanted, but it should allow you to do more than you were previously able to.

5: Integrate all your systems

I suppose this wouldn’t be complete if I didn’t suggest that you integrate all your systems.  It’s highly unlikely this will work out for you, but I think I should mention it.

Why won’t it work out?  Well none of your systems are ever going to match up.  However if you can make it so that you are only using one system, then you may be in a better world.  The difficulty of this is that very few companies make a tool that can do all these things and many companies have finance tools that are too indoctrinated in the company culture to be able to just change.  Especially if you sell through many channels.
My advice? Your best value for money is through going through steps 1 – 4 above.  Most of them will give you quick, cheap wins.  If you’re Amazon or eBay you can go for option 5, but for the rest of you it probably isn’t worth it.


Why A/B testing?

A/B testing, for those of you who aren’t sure, is when you offer the users of your website an option between Choice A and Choice B, and use their preference to help you decide which to use. Sounds pretty simple, right?
Well, in some ways, it is. Often, even a simple change can make a drastic difference in how your customers perceive your website, and can lead to dramatic changes in your conversion rates – one of the most important metrics you should be keeping track of. However, one can’t always guarantee that A/B testing will yield such amazing results instantly, so it’s important that we keep a few things in mind when setting up your test:

Things to keep in mind:

  • A Dramatic Change Is Not Always Necessary.
  • Many Small Changes Can Be Made For Tailored Results.
  • Don’t Rush The Test – Give It Time To Be Statistically Significant
  • Don’t Test For Too Long, Either.

If you keep these simple ideas in mind, your A/B test should be significant, without forcing you to completely redesign your page. After all, if you already worked hard on your User Experience, then you’re not necessarily going to need to rip out the whole thing and start over. Instead, you should focus on improving little bits at a time, and remember -

Small Changes Add Up

Once you have chosen a page, decide on one element that you would like to use as your testing element. Be bold when it comes to your testing element.

- Mark Thompson, Pro Blog Design.

One of the things that every small business owner should remember is that, although change takes time, not every change has to be huge. You may have a great product, you may have a great website, but what separates the eternal small business owner from that future tycoon is the drive to improve. The nice thing about A/B testing is that, although it’s simple to set up, it allows you that most coveted of all business perspectives: the ability to “read the minds” of your customers.

But while A/B testing gives you access to this holy grail of business treasures, it is best to remember that the consumer pool is best absorbed in little bits. After all, if your product is already selling, then it makes little sense to completely rewrite your page copy. The old adage “If it isn’t broken, don’t fix it” is alive and well in eBusiness.  But, consider where we would be today if people took every chance they had to improve their good ideas. Instead of approaching changes as grand and sweeping, let the ease of A/B testing allow you to focus on making small changes – adding a word or two, tightening up your copy a bit, adding images, better describing features, or all of the above. Figure out which ones your customers really like, and let that drive sales. Bits and pieces add up. Knowing whether your  customers prefer a streamlined interface to a page detailing every piece of information about your product or service, can lead to drastic increases. Often times, your personal information may disagree with what intuition, or the experts, say. Additionally,  small changes can give you a wealth of information, especially when bolstered by well defined metrics.

It is important to remember, though, that customers can become burned out on A/B testing. While, generally,  people like companies that are trying to improve their user experience, it is easy to focus so much on pleasing the “committee”,  that you lose your own personal voice and style. There is something insincere about a website that sacrifices its own point of view for only what the consumer wants. You can’t please everyone, and it takes a certain level of confidence to have that defiant point of view. In the end, your personal tone is what separates your site and product from the imitators and what has come before. So remember: A/B testing gives you a peek into the customers mind, but they’ve come to buy a piece of your mind. Use testing wisely, use it well, but don’t let the customers become burned out on it.

Now, go forth and test. You may be surprised what you discover lurking in your customers’ heads.

I’ve had a lot of conversations lately about the strategy of group buying sites (or daily deals, flash sales, etc.). Groupon is the leader in this space…so much the word is becoming a verb. The questions I often hear are: How do you know if Groupon (and group buying deals) are right for a type of business? What are the factors that make Groupon a profitable strategy? How do you evaluate and analyze the profitability of Groupon?

Already there are a lot of competitors with Groupon, and more several that are headed toward even more niche group buying capabilities, focused by interest, small city, or people groups. The group buying strategy will continue, and so will the conversation about this. But the model of giving a significant (50%+) discount on goods and services has its dangers. So it piqued my curiosity to analyze this from an economic perspective.

On the plus side, this is a pay-for-performance approach to customer acquisitions. And it’s a sudden and (mostly) predictable burst of new customers and revenue.

On the cautionary side, you’re paying for that acquisition with negative margin. Do business owners really know (or at least rationally evaluate) the complete profitability of these customers?  If I were doing this, I’d look as much as possible at the total economic impact, as there are some overlooked aspects to this type of promotion.

A fascinating study on Groupon effectiveness by Utpal Dholakia of Rice University cited that 66% of small business owners report Groupon to be profitable. In discussing this with him, that figure is a self-reported, which is valud to understand how the owners think about the outcome. It is natural for an owner to believe and report they made a profitable decision. But this is not to say that 66% of Groupon promotions actually are profitable. I would assert most small business owners were not instrumented or had taken the time to fully analyze profitability.

Utpal and I agree that the analysis for full profitability may not be possible for the average small business owner. Epsecially at the detail I’m about to outline. However, if I were running a business, I’d at least want to logically think through the assumptions of profitability and measure what I could…otherwise, I would be headed down a slippery slope.

This kind of customer acquisition can become a ‘drug’ to a business looking for revenue, and yet the total P&L impact may not be understood. Groupon reports that 95% of businesses would run Groupon again, though Utpal’s study suggests it’s more like 68%. Both are self reported figures. How many actually DID use Groupon again? I digress.

Let’s use a fictitious example to walk through what the “ideal” analysis would include.

Groupon Assume you’re the owner of a spa salon and you offer a coupon of $50 for $100 of spa services (to make it easy), and you sell 1,000 of them.

You get $25 from each sale (because typically 50% of the $50 goes to the group buying site). If you had 50% margin on the $100 list price, then you’re losing $25 on each deal and 1,000 of these coupons is costing you $25,000 in negative margin.

On the plus side you’ve acquired 1,000 new customers. However, how many of those actually ‘new’? This is the first key assumption and the maturity and visibility of your company will be important in determining this value. Let’s assume 20% of those who received the coupon would’ve bought at full price. That’s 200 customers that would’ve given you $10k in margin, but instead cost you $5,000. That’s a $15k net swing.

The remaining customers are new, 800 customers that cost you $20k in negative margin But, how many will buy again at full price over the year? This is another key assumption and the type of business you have an the kind of service or product you provide have impact on the lifetime value calculation. For this exercise let’s assume 20% of the 800 new customers will come back and spend $100 in services again three more times in the year. That’s 160 customers driving $150/yr in margin (3x $50 margin) = $24k in margin.

Here’s the margin math so far:

200 existing customers in lost margin = -$15k

640 customers who won’t come back = -$16k

160 customers with 3x 1yr full margin value = +$24k

It’s unprofitable soi far. Ah, but we’re not done!

How many of the 1,000 customers never redeemed the coupon? Let’s assume 10% don’t execute on the coupon before it expires. That’s 100 coupon purchases where you get $25 each with no cost of goods, $2,500 in positive margin.

Now, how many of the 900 customers who DO redeem buy something else when they turn in the coupon? Let’s assume 30% of those customers spend 30% more. that’s 180 customers spending $30 in full margin ($15) = $2,700 positive margin.

And, what’s the brand recognition worth of the campaign itself? This is the most difficult to measure and understand. Yet it’s probably the assumption Groupon wants you to believe in the most, which is one of the reasons they invest in great copywriting for their offers. There are a lot of assumptions to think through on this…how many are seeing the promotion, how well is your company marketed, how many of the audience already knew about your business, what’s the acquisition opportunity of this kind of awareness-building, etc.? For the sake of this exercise, let’s just assume that 25,000 people see this promotion and 1% of that audience will visit you at full price, assuming the same $100 of service they purchase 4x a year. That’s 250 new customers spending $400 ($200 in margin) = $50,000 in margin.

Ok, so let’s net out the total economic impact:

200 existing customers in lost margin = -$15k

640 coupon customers who won’t come back = -$16k

160 customers with 3x 1yr full margin value = +$24k

100 coupons not redeemed = $2,500

180 customers buy more on site = $2,700

250 new full margin customers from campaign awareness = $50,000

Total margin impact for year: $48,200

Now, there are a ton of assumptions in this exercise you can argue. I just made these numbers up.  If the awareness didn’t bring any customers in after the coupon, the example is not profitable. The point is to illustrate the factors to think through and debate with yourself.

Also, every business is different, every group buying site campaign is different, and you could do it at different times of the year which would all effect the economic outcome. The one-year value of doing a group buy coupon could be negative margin as much as it could be positive margin in this exercise.

For example, when Groupon did the Gap promotion, perhaps the % of customers that would’ve bought from them anyway is much higher.  Everyone is aware of Gap, so I doubt they got as much upside on the awareness building from the campaign. If anything, they may have told customers that it’s possible to get a better deal at Gap if you look for a coupon or wait, so perhaps they lost a higher percentage of full margin customers. The turnout for Gap may have been very bad from a one-year margin calculation and a brand impact. Or, perhaps the majority of people who redeemed coupon spent twice as much at full margin and therefore they didn’t lose money. Though I still question the incremental lifetime value opportunity from these customers. In my experience coupon users are discount shoppers. The thrill is in getting the best possible deal, so money spent beyond the coupon is not a deal.

The point is there are a lot of factors that go into the determination if this is a good strategy for one business vs. another.  Here are some factors I would consider if I were a small business considering if a group buying strategy was worthwhile:

  1. Awareness – if your company already has high awareness in your market then the awareness building benefit of the group buying campaign is less positive. Further, if you have high awareness, than the % of existing customers who buy with a coupon could be high.
  2. Repeat Business – Do you have a business that has repeat customers? 50% off a vehicle registration service is not as good as 50% off a haircut. If you have repeat business, then you have a higher likelihood (and of calculating) for lifetime value of a customer acquired.
  3. Differentiation – How differentiated is your business, product and/or service. If you’re a burger joint as good as many others, than customers may take you up on your coupon, but go to the next burger joint with a coupon next time.
  4. Word of Mouth – Related to repeat business, what is the likelihood that someone who experiences your business at a discount will be delighted, and how word-of-mouth-worthy is your experience. Interestingly, Certain types of businesses or products can have a higher word of mouth quotient than others. A zip line company has more WOM than a nail salon. And customer experience matters. Uptal’s research reports that if employees were happy with the incoming customers, the owner was more likely to believe the promotion was profitable. Things like group-buying sites will require businesses to create better products and services, otherwise you will lose the LTV benefits of acquiring customers. See my Mashable article.
  5. Upsells – What’s the likelihood people who come in with a coupon will buy more at full price? The type of product and service will make a difference, as much as how upsells are executed. According to Utpal’s study,
  6. Breakage – what is the likelihood people will not execute on the coupon? Are you far away; is the dollar amount small, is scheduling involved?
  7. Data – How much data can you capture about the customer? That enables you to remarket (assuming you do so) and raise the return business assumption. And what is your ability to capture and analyze the measurements above? This doesn’t change the outcome, but it does better inform you if you should do it again. Otherwise, it may be too tempting to do something like this again that could put you out of business.

Of course, I didn’t calculate your time and opportunity cost to evaluate, execute and analyze all this to determine if you should do a group discount. Add that to the mix as well!

Again, I don’t assume small business owners can or will go to this level of detail in analysis. The point is to consider the characteristics above, to go through a rational decision making process, and then measure what you can to determine how to use group buying.

Guest post by Major Hayden.

I’m finally delivering on a promise to my readers which I made a few months ago. I’ve written a guide on how to host a web application redundantly in a cloud environment. While it’s still a bit of a rough draft, it should be a good starting point for those who haven’t worked in virtualized environments before. Also, it may show some of the more experienced systems administrators a new way to do things.

As always, if you find anything in the guide that needs improvement, I’m all ears.
Redundant Cloud Hosting Guide
The purpose of this guide is to answer one of the questions I receive most often:

How do I host my web applications in the cloud in a way that is redundant but also inexpensive?

Before you begin reading the guide, try to keep the following things in mind:

  • Try to understand what an application is doing before blindly configuring it as the guide states. This helps in two ways: it allows you to begin thinking about ways you can improve your configuration for your specific needs and it will give you tools to fix things when they break later.
  • Stay lean. There may be some portions of the guide which may not apply to your application’s needs. Instead of wasting your time on additional daemons that you don’t need, skip over any parts of the guide that don’t apply to your specific application. On the other hand, if you find that your application needs more functionality than this guide provides, be sure to add in extra functionality carefully. See the previous bullet point to understand what I’m talking about.
  • This is not the only way to configure a redundant cloud environment. This guide covers the configuration that I like best. If you don’t like a particular daemon or Linux distribution mentioned in the guide, use what you’re most comfortable with or what you prefer.
  • Cloud is what you make of it. Don’t be afraid to forge your own path.
  • Give me feedback. If you spot something that’s incorrect, or if you find a more efficient way to handle a particular problem, let me know! I’ll be glad to consider it for the guide and you’ll receive proper attribution.

With that out of the way, let’s begin the guide.

 


The high-level overview
To get an idea of the end result, review the diagram shown below:

 

Redundant cloud hosting configurationMy redundant cloud hosting configuration includes two load balancers, two web nodes, and two database/caching nodes. They all have interfaces on public and private networks.

There are three main service groups that I need to host my applications:

  • Load balancing layer: Two needs are fulfilled at this layer – the distribution of load as well as redirection of traffic away from problematic web nodes.
  • Web service layer: As you could imagine, this layer is the workhorse of the entire configuration. This is where web content is served and where web content is stored in a clustered filesystem.
  • Database/caching layer: Without this layer, the configuration would grind to a halt. The applications running on the web services layer depend on this layer for rapid storage and retrieval of information.

Platform requirements
In order to follow this guide, you’ll need the following:

  • Stable Linux distribution – pick whichever one you prefer, but I’ll be using Fedora
  • Six virtual machines – anything less than six will get a bit tricky and it reduces your redundancy
  • Public and private network interfaces on each virtual machine – not required, but it’s highly recommended
  • One extra IP address – this will be your virtual IP address for load balancing (you will need more if you’re hosting multiple sites with SSL, unless you want to use SNI)
  • Ability to share an IP between multiple virtual machines – this will be a requirement for LVS-TUN (if you can’t share IP’s, you can try using LVS-NAT, but I wouldn’t recommend it)
  • Kernel modules – you’ll need a few kernel modules, or the ability to compile and use them with your running kernel
  • Linux kernel 2.6.27 or later – there are some great performance improvements for virtual machines and the fuse module in these kernels (not a strict requirement, but highly recommended)

Step by step
I’ve broken the guide up into functional pieces to allow you to build your configuration and test it along the way. Click on the title of each step to see detailed instructions, diagrams and explanations:

 


What’s the total cost?
Right now, I’m hosting this configuration with Slicehost with the following setup:

 

  • load balancers: two 256MB instances (2 x $20/month)
  • web nodes: two 1024MB instances (2 x $70/month)
  • database nodes: two 512MB instances (2 x $38/month)

That adds up to $256 per month for the entire configuration at Slicehost. That price also includes 2.1TB of public bandwidth (since the bandwidth is pooled between instances). The only large consumers of bandwidth are the web nodes since they send out a lot of traffic. The load balancers simply receive requests on the public interface and shuttle them to the web nodes over the private network. The database servers would only talk to the public network for package updates.

If you wanted to host the same configuration with Rackspace’s Cloud Servers, you could do it for as little as $153.30 per month, but your bandwidth would be billed at the utility rates. For low traffic sites, this may be the better-priced option.

Originally posted here: http://rackerhacker.com/redundant-cloud-hosting-configuration-guide/

© Major Hayden 2010.

This post was originally posted on August 11, 2010 on Major’s blog, Racker Hacker.

Major Hayden is a Linux Systems Engineer for Rackspace in San Antonio. He works with the Cloud Servers and Slicehost virtualization products. Major’s primary focus is on base image maintenance, kernel customization and tactical optimization solutions. He also maintains multiple blogs and a MySQL optimization script called mysqltuner. Outside of Rackspace, Major enjoys contributing to the open source community, running, and taking care of his chinchillas.

Below, the top ten online display advertising networks in the U.S. by reach, according to data from comScore.

The measurement firm estimates Yahoo’s network continues to reach more U.S. users than any other, with 85.9 percent audience penetration. AOL Advertising and Google’s Ad Network follow it, with 85.7 and 82.7 percent, respectively, followed by ValueClick and Turn Media.

Networks operated by 24/7 Real Media, AdBrite, Collective Network, Specific Media, and Microsoft round out the top ten.

adnets