You’ve just posted a killer blog entry and submitted the link to digg. You get a hundred or so diggs in the first hour, and the next thing you know, you’re on the front page. A couple hundred diggs later, you get one of the various dreaded messages that your site is either dead or dying (“Server cannot be reached”, “Service Temporarily Unavailable”, or the unthinkable “Account Suspended!”).

“How could I have prevented this?”, you ask yourself.

There have been many articles on this topic offering various suggestions such as toying with the webserver and database settings, using a form of query/content caching, and ensuring you use good hosting. All are good suggestions, but what if you’re using a blogging service or shared hosting? The following list contains some very simple ways to help alleviate some of the issues caused by the digg effect.

1. Monitor the number of diggs: It’s not as if you wouldn’t do this anyway, but keeping an eye on the number of diggs and if/when you get front paged helps you get a jump on the traffic spike.
2. Talk to your hosting provider: Giving your hosting company a heads-up on what’s going on can save them some headaches (especially if you’re on shared hosting) and possibly save you some $$ on your hosting bill.
3. Turn off comments, widgets, etc: Most diggers don’t comment on dugg pages anyway, so one less trip to the database can make a big difference. Have a rating widget you can disable? Do it. Try to keep the number of queries to the database to a minimum.
4. Don’t use images or other media: Content is king, right? Unless you’re specifically posting an image or some other media, don’t include additional images or media. Each one is another request that the server has to handle and another chunk out of your bandwidth cap.
5. Post a static copy: Some blogging tools will create a static copy of your post. A static page uses much less server resources to serve than a dynamic one. If you don’t have that option, create a static copy and redirect to it. If things are still slow, post a “print friendly” text only version that removes all the excess chrome from your site.
6. Redirect to a copy on a caching service There are a couple of caching services that are great for situations like this: DuggMirror and CoralCache. They’ll cache a copy of your page and have the infrastructure to handle very high loads. It’s a good idea to follow their instructions and populate a cached copy as soon as you submit to digg. With the more popular stories, diggers will refer to the cached copy when the dugg page gets slow, but often the cached copy is too late and grabs a copy of one of the dreaded “dead site” messages. Better yet, why not create a cached version using CoralCache and submit the link to the cached version to digg.

I can’t guarantee that doing all these things will keep your site from faltering to the almighty digg, but they’ll certainly help keep it alive as long as possible. Happy digging!

I have completed a new page entitled “Essential Reads” — you’ll find a link to it located on the right hand side just below the PageRank Decoder. I think you may find the page useful. There’s three sections to the page:

1. Find All Post by Author
The top part allows you to find all articles by a given author of Search-This. This was actually fun for me to develop because I got a chance to use both Flash and Ajax. If you mouse over a team member it displays their user name and real name and then if you click it, it retrieves all articles by that user. If you don’t have Flash you can still use the dropdown menu.

2. Essential Post
This next section shows some post that I think all would benefit from reading. This would also be a good time to subscribe to the Search-This feed. We now allow you to subscribe via e-mail too, so please sign up today.

3. Top Ten List
The last section simply shows the last ten people to comment on Search-This. This can be handy for trying to keep up with all the comments throughout the website. Below the last ten comments list is another list of the top ten most commented-on posts.

I hope you find the new Essentials page useful. If you can think of something else you would like to see included on the page, please let me know.

Thanks,
Mark

“Know Thy Enemy” – The Art of War, by Sun Tzu

“If you know the enemy and know yourself, you need not fear the result of a hundred battles. If you know yourself but not your enemy, for every victory gained you will also suffer a defeat. If you know neither the enemy nor yourself, you will succumb in every battle.”

When you run a business, it’s always a good idea to know who your competition is as this is what really keeps businesses evolving. Recently, our company was recognized as an Authorize.Net certified developer. As an aside, Authorize.Net is the most popular payment gateway which acts like your website’s credit card terminal. Back to my story… since this has happened, we’ve had a ton of phone calls asking us to integrate Authorize.Net into an already existing website. This increase in traffic and sales has been extremely welcome, but after a couple months, we realized that we were landing every single lead that came our way. And when we told them our price, people immediately said , “YES!” So, we wondered if our prices were too low.

(more…)

Sometimes I need to pull files from one server to another so I need to use secure copy. This is really simple to do with SSH, but I always seem to forget the command when I need it. SSH allows Unix users to secure their terminal and file transfer connections. Now let’s look at the code:

Once you are where you want to be, the command is:

scp filename.ext ses5909@123.123.123.123:

First type in scp (the secure copy command), then the filename, followed by your username on the other server, the @ sign, and the IP or host name of the other server. The part I always forget is the colon (:) and you need this or it will not work. Once you’ve executed this, you will be prompted for your password on the other server. Once verified things should transfer to your user directory on the other server.

If you need to copy an entire directory be sure to include the -r and you’re set

scp -r directory ses5909@123.123.123.123:

Hopefully this helps someone else out! Do you have any basic commands that you always forget?

If you hang out on enough web development forums, mailing lists, or chatrooms, you’ll eventually get asked this question:

What language should I use to program my next project?

As someone who has been on a few of said forums, mailing lists, and chatrooms for a while now, I hate seeing this question. Yes, it comes around a lot, but repetitiveness isn’t what bugs me about it. What bothers me is that there’s usually very little background information attached to that question, which makes it as open-ended as asking what mode of transportation you should take to get to the store. I can tell you anything from a skateboard to a jet and either vehicle can get you to the store, but to give you the best answer I need to know how far you have to travel, what you plan on buying, and how much money you have, among other things. The same holds true for the programming language question, so let’s look at some factors that can help you choose which language to use on your next project.

(more…)

On my server, I have several domains and I have corresponding email addresses for a few of them, for example: ses5909@mydomain.com, ses5909@anotherdomain.com, etc. I have come to find that in Plesk if I have a domain that has an email user named ses5909 and I need to make the same username on another domain, they are not allowed to share the same domain. So if I give ses5909@mydomain.com the password: supersneakypw, I cannot give that password to ses5909@anotherdomain, or anyone else that has the same username. This is pretty lame if you ask me. The reason for this is that mail is not handled on the domain level, but rather the username level. That is a whole other issue though.

Well today I needed to get into my mail and I wasn’t at the computer that I run my mail client on so I went to my webmail account. I couldn’t remember the password for the life of me. I tried a few that I commonly use, but no luck. So, I needed to try to find my password. I could have reset it, but I’ve resorted to this in the past and I am just tired of doing that. I needed to find a way to retrieve my password.

So, I logged into mysql as the root admin. I looked through the databases and noticed there was a psa db which I am assuming stands for Plesk Server Administration. I immediately looked for email and finally saw mail. I saw all of the usernames and a quick join with the domains table showed me which column belonged to which domain.

SELECT mail.mail_name, domains.name from mail LEFT OUTER JOIN domains ON domains.id = mail.dom_id;

But, now I needed to find the passwords for these. After some more hunting, I found the accounts table where passwords are stored in plain text. If only I didn’t have 200 to look through. So a query was in order:

SELECT accounts.id, mail.mail_name, accounts.password, domains.name FROM domains LEFT JOIN mail ON domains.id = mail.dom_id LEFT JOIN accounts ON mail.account_id = accounts.id

Problem solved

What is the PageRank Decoder?
The PageRank Decoder allows you to visually see how Google factors PageRank based off of link structure. Be sure to view the Demo first.

What’s it good for?
You can examine different link topologies and see how PageRank is passed amongst pages. This allows you to see if your websites link structure is as optimized as possible.

How should I start?
I would recommend that you start by adding a bunch of pages to the canvas and then start interlinking them – then hit the calculate button and hover over the toolbars for their PR value. This allows you to see how different link topologies pass PR.

Later, for fun you can start looking up pages by entering in their URL in the textbox.

How does it work?
You should start by reading the article, “Google’s PageRank Explained.” Once you have a firm understanding of the article you will have a better understanding of how the PageRank Decoder is working.

Is this thing really accurate?
Yes and no?

To calculate the PageRank for a page, we use Google’s original formula:

PR(A) = (1-d) + d(PR(t1)/C(t1) + … + PR(tn)/C(tn))

If the above code looks foreign to you and you have never seen this formula before, then start by reading, “Google’s PageRank Explained”.

It’s accurate if all pages start with a PageRank of 0. If you looked up a page or set a page’s PR at something other than 0 it can mess up the formula. This is because that page now starts off with a PR other than 0 and that then gets compounded in the formula.

Google doesn’t do this; they start all pages at 0 (or rather 0.15) and then run their formula for each page factoring all incoming links. Google’s database exceeds a million pages and for each page they look at its incoming links, some pages may have well over 1000 incoming links. It takes Google somewhere around a month to update all pages.

The PageRank Decoder is not working with a database to track incoming links for a page; rather it just looks at what’s on the canvas. So instead the PageRank Decoder allows you to artificially set a page’s PR.

I thought a ranking can’t go beyond 10?
Remember, there’s a difference between PageRank score and the toolbar-shown PageRank. The toolbar has a max score of 10. I don’t know Google’s scale when setting the PageRank in the toolbar so it gets inflated quickly.