ServerLoft: a big NO.

As I wrote in a previous post I used to use ServerLoft’s services. After 1 year I cancelled my servers and I decided to share with you my experience so you may choose not to repeat my mistakes :-)

Here’s why I wouldn’t advise you to (ever) use ServerLoft:

1) Not the best offer in terms of value for money
Though their offer seems to be one of the best available, and it is in terms of money, on the long run you’ll end up losing more money.
That’s because:

  1. Their prices do NOT include VAT and depending on your country you’ll discover that you have to pay 19% more
  2. There are better offers for less money (and, I admit, less traffic included)
    If you don’t need the monthly 5-10-15TB of traffic I would recommend you using Hetzner. Their servers are one of the bests and cheapest. And their price include VAT so you may end up, depending on your country, paying less,

2) Hardware quality is sometimes poor
After using for 1 year ServerLoft I figured out that:

  1. You shouldn’t try to reboot your server if that’s possible. It usually takes a server about 25 minutes to reboot. Don’t ask me why and it is not dependend of the type (hard or soft) of reboot you ask for: it just takes 25 minutes to reboot!
  2. Their advertised hardware RAID1 is useless (see my real story below) and poorly documented.

3) RAID 1: a joke
This is the real story of my experience with what RAID1 means for ServerLoft:

  1. You won’t get any meaningful support for their (advertised) 24/7 support. I mean, let’s suppose that I would need to know the status of my RAID1 or I would like to know if I configured/tweaked it properly. Here’s the kind of arrogant answer I got from them on my request to just provide me with some links/documentation:
    it seems we cannot help here. Serverloft is a service for very professional guys. Those guys should be able to setup a network, a raid-system or somethink else. We do not offer a kind of technical support, we just help you if your hardware died and have to be replaces.
  2. After using their services for 1 year I noticed that the sites on one of my servers started to behave strangely. I logged in my server and tried to figure out what’s wrong.
    After some investigations I’ve figured out that the filesystem is in Read Only state.  Reading about this on the net I started to believe that a HDD may be broken.
    I know that it is STRANGE (to put it mildly) that a “state of the art” hardware RAID 1 will set the OS as Read Only when a HDD is broken but I tried to look at the bright side: I haven’t lost any data, I just have to figure out what HDD is broken, contact ServerLoft support and they’ll replace it in a matter of maximum 4 hours as advertised on their site.
    Well…here comes the funny part: I run their automated hardware test just to learn that my system was in perfect shape (see here)

    After receiving this report I wanted to boot again and investigate further. I tried to reboot the server. And tried …and tried. After 3 hours I sent the guys a message asking them for help. After 12 hours not receiving any answer I wrote them again. Then about some 18 hours I called them…and called them again after 24h …and again after some 36 hours…
    No, this is not a joke, it happened to me: I had to write them some 4 messages and call them 3 times to get, after about 48hours this answer:
    The hardware reboot is working if request over the PowerPanel but your system could not be booted because the bios did look for bootable system in the wrong device order. We have fixed it and your system is now up and running again.

    We also think the wrong drive was marked as failed, so now we have one damaged harddrive with badblocks (and complete raid) and one functional drive with incomplete raid. You should check it and if we are right we ask you to backup your data (maybe in recovery system) if needed. After that please tell us to exchange faulty drive and create new raid. Reinstalling the operating system is neccessary and will result in a complete data loss.

    So, after waiting for 48h+ this is the kind of answer their support can provide. No “apologize”, no “sorry” no nothing. After waiting for their support for 48+ hours I realized that I paid 1 year for nothing: their “RAID” is actually useless and that I’ll have to backup my data and reinstall my server. They cannot help me more.
    I tried to discuss with the sales guys, ask them whether they feel that this is normal and if they don’t feel like they should at least apologize and give me a discount…something…anything!
    I wrote them 2 messages and never get an answer back from them.
    They really don’t give a shit. They just sent me the bill as usual.

4) Absolutely useless customer support

  1. Their advertised 24/7 support is actually a “get a meaningful answer after 48h and that involves reinstalling your system”.
  2. They aren’t able to provide you with any support or useful information at all.
    And when I say that I think of Hetzner, another German dedicated hosting provider.
    They don’t advertise that they have 24/7 support but it never took them more than 2h during business hours to answer me and my last experience with them was wonderful: I asked them to replace my server’s RAM because I suspected it may be broken. That was 30 minutes before their support hours ended.
    I got an answer from them after 1 hour (so NOT during their support hours): their technician  informed me that he replaced the RAM and he also advised me to replace my server’s network card driver as the default one had some problems. And he not only told me that but he also provided me with a link to a step-by-step documentation (in English, of course) on how to upgrade my software driver.
    No comments.

So, that’s my experience with Serverloft. I’ll end by saying that I decided to cancel my servers at Serverloft and move to Hetzner: it took me less time to order a new server, install it and move all my sites to the new server than it took me to get any meaningul support from Serverloft.
No, really, I’m not joking!

Debian 5.0 + Apache 2.2.9 + PHP 5.2.6-1 and MySQL 5.0.51 @ Hetzner

I recently started using a brand new server from Hetzner.de and as I spent some time finding the best configuration and settings I thought it may be a good thing to share all the empiric knowledge I gathered so that others save time and frustration. Hopefully this will also bring me some points in Heaven but that’s another thing :P

First of all a few words about Hetzner: I work with them since 3 years now and I can say that they have one of the best (if not THE best) offers available: 8-CPU Intel® Core™ i7-920 Quad Core with 8GB RAM, RAID 1 750GB HDDs and 2TB of traffic at 100Mb network speed.
They don’t have 24/7 support but the support they have is pretty good and they were always quick in replacing HDDs/RAM when I needed.

That’s being said here’s a step by step configuration for optimal performance. From my experience a server configured in this fashion can easily stand about 200K uniques/day with MySQL serving over 2K queryes/second.
Please note that this is the exact custom configuration that I use, you can choose to use it yourself or you can pick just what you want out of it.

Step 1 - create some directories and link
mkdir /usr/local/install
mkdir /usr/local/important
mkdir /usr/local/z_delete_later
mkdir /backup
mkdir /usr/local/important/apache_docs
ln -s /usr/local/important/apache_docs/ /web
mkdir /usr/local/important/mysql_data
mkdir /usr/local/important/apache_docs/phpmyadmin
ln -s /usr/local/important/apache_docs/ /web
mkdir /etc/firewall

Step 2 - install packages
This step is supposed to be 99% automated, just copy and paste the commands below in a file, chmod +x that_file and then run ./that_file
You should just a couple of time choose some configurations and that’s it

apt-get -y install tofrodos
apt-get -y install catdoc
apt-get -y install sudo
apt-get -y install whois
apt-get -y install sysstat
apt-get -y install lsof
apt-get -y install gawk
apt-get -y install dstat
apt-get -y install unzip
apt-get -y install mc
apt-get -y install postfix
apt-get -y intall ftp
atp-get install host
apt-get -y install hdparm
apt-get -y install links
apt-get -y install lynx
apt-get -y install bzip2
apt-get -y install gcc
apt-get -y install make
apt-get -y install g++
apt-get -y install file
apt-get -y install ncurses-term
apt-get -y install  libcap-bin
apt-get -y install  libcap-dev
apt-get -y install libncurses5-dev
apt-get -y install zlib-bin
apt-get -y install zlib1g-dev
apt-get -y install zlibc
apt-get -y install libxml++2.6-2
apt-get -y install libxml++2.6-dev
apt-get -y install libbz2-dev
apt-get -y install libiconv-hook-dev
apt-get -y install hicolor-icon-theme
apt-get -y install gnulib
apt-get -y install perl
apt-get -y install libperl-dev
apt-get -y install mime-support
apt-get -y install telnet
apt-get -y install iptraf
apt-get -y install autoconf
apt-get -y install psmisc
apt-get -y install sshfs
apt-get -y install afuse
apt-get -y install fusesmb
apt-get -y install libfuse-dev
apt-get -y install host
apt-get -y install libevent-dev
apt-get -y install ftp
apt-get -y install gawk
apt-get -y install gawk
apt-get -y install php-pear
apt-get -y install mysql-client
apt-get -y install kdelibs4c2a
apt-get -y install rar
apt-get -y install unrar
apt-get -y install apache2
apt-get -y install apachetop
apt-get -y install libapache2-mod-php5
apt-get -y install php5
apt-get -y install php5-gd
apt-get -y install php5-mcrypt
apt-get -y install php5-curl
apt-get -y install php-db
apt-get -y install php5-mysql
apt-get -y install phpmyadmin
apt-get -y install mytop
apt-get -y install mysql-server
apt-get -y install mysql-client
apt-get -y install backup-manager
apt-get -y install libgd-dev
apt-get -y install php5-dev
apt-get -y install nmap
apt-get -y clamav

apt-get -y curl
useradd apache
ln -s  /usr/bin/curl /usr/local/bin/curl
ln -s /usr/bin/php /usr/local/bin/php

That’s all, the last 2 lines are just for script compatibility purposes.

Advertisement:

Step 3 - Fix the SSL certificate
wget http://1mum.geek-tools.org/wp-content/uploads/2009/11/apache2-ssltar.gz
tar zxv apache2-ssltar.gz
sudo mv ssleay.cnf /usr/share/apache2/
sudo mv apache2-ssl-certificate /usr/sbin/
mkdir /etc/apache2/ssl

then run
apache2-ssl-certificate

then

cd /etc/apache2/
Here you will see two folders; sites-enabled and sites-available. First make a copy of the current available site and name that “ssl”. cp sites-available/default sites-available/ssl Then make a symbolic link to your new configuration to enable the properties. ln -s /etc/apache2/sites-available/ssl /etc/apache2/sites-enabled/000-ssl

Now you must configure the “ssl” host.
Relplace
NameVirtualHost *
<Virtualhost *=”">
ServerAdmin webmaster@localhost

DocumentRoot /var/www/</virtualhost>

with
NameVirtualHost *:443
<Virtualhost *:443>
ServerAdmin webmaster@localhost
DocumentRoot /var/www/
LoadModule ssl_module /usr/lib/apache2/modules/mod_ssl.so
SSLCertificateFile /etc/apache2/ssl/apache.pem

SSLEngine On

and finally configure apache to listen on port 443 (https) by editing /etc/apache2/ports.conf.
Restart Apache  with /etc/init.d/apache2 restart and you should be done with this part.

Are you still there? Admit it, it was piece of cake till now and we’re half way done :-)
So let’s move to

Step 4 - different twaekings
# mv /var/www in /var/www2
create a symbolic link from /var/www to /usr/local/important/apache_docs ( ln -s /usr/local/important/apache_docs/ /var/www )
create a symbolic link from /usr/share/phpmyadmin to

/usr/local/important/apache_docs/phpmyadmin_new
# mkdir /etc/firewall then #vi firewall
and configure your firewall as you wish.

Next step is to configure /etc/rc.local, here’s my example
/bin/echo “1073741824″ > /proc/sys/kernel/shmmax
/etc/firewall/firewall &
/sbin/hdparm -a 2048 /dev/sda
/etc/init.d/postfix start
Explanation: first line is for eaccelerator (will discuss it a bit later), firewall is clear, hdparm is for speeding up the HDD a bit so it reaches the optimium performance and postfix is started.

change the server name by editing
#vi /etc/hostname

then
#vi /etc/profile
and add
alias l=’ls -al’
alias ll=’ls -al|more’
alias c=’clear’
export EDITOR=/usr/bin/vim

#vi /etc/vim/vimrc
and uncomment the line with syntax on

#cd /usr/local/important/apache_docs
#ln -s /usr/share/phpmyadmin/ PHPMyAdmin

Step 5 - MySQL tweaking
There are many options for /etc/mysql/my.cnf but I won’t describe them here (if anyone interested just drop a message). I’ll just explain some configurations that I use so I goup all my important data in /usr/local/important

#mv /var/lib/mysql/* in /usr/local/important/mysql_data
#cd /var/lib
#ln -s /usr/local/important/mysql_data/ mysql
#chown mysql:mysql mysql_data/ -R
#chmod 700 mysql_data/ -R
Make symbolic link from /var/lib/mysql to /usr/local/important/mysql_data

Step 6 - Activate mod_rewrite
For this I got inspiration from http://www.lavluda.com/2007/07/15/how-to-enable-mod_rewrite-in-apache22-debian/

Basically what you need is:
#find / -name “mod_rewrite.so”
and result should be /usr/lib/apache2/modules/mod_rewrite.so
In /etc/apache2/mods-available you run#
ln -s /etc/apache2/mods-available/rewrite.load rewrite.load
then make sure the module is loaded upon apache restar
# ln -s /etc/apache2/mods-available/rewrite.load /etc/apache2/mods-enabled/rewrite.load

Step 7 - Configure Apache for high performance
edit /etc/apache2/apache2.conf
Below are the settings that I choose to edit, you could add more if you want
- ServerTokens ProductOnly
- Timeout 10
- KeepAlive On
- MaxKeepAliveRequests 500
- KeepAliveTimeout 5
- <IfModule mpm_prefork_module>
StartServers         50
MinSpareServers      25
MaxSpareServers      50
ServerLimit           500
MaxClients          490
MaxRequestsPerChild   100
</IfModule>
- <IfModule mpm_worker_module>
StartServers          20
MaxClients          350
MinSpareThreads      25
MaxSpareThreads      75
ThreadsPerChild      25
MaxRequestsPerChild   0
</IfModule>

Step 8 - Some PHP tweaking
Below are the items I usually change
#vi /etc/php5/apache2/php.ini
- expose_php = Off
- memory_limit = 128M
- register_globals = On
- dezactiveaza partea de erori (display_errors = Off)
- post_max_size = 64M
- upload_max_filesize = 64M

Step 9 - Compile and configure eAccelerator
I always use on my servers a PHP accelerator, whether that’s it APC or eAccelerator doesn’t matter that much (though I prefer eAccelerator). I recommend you do the same, it will dramatically improve your web server’s performances.

Download release 0.9.5.3 from here http://bart.eaccelerator.net/source/0.9.5.3/
At the moment when I write this doc I also tried 0.9.6-rc1 but I definitely not recommend it: it is not yet stable and caused me a lot of problems with apache becoming unusable after a couple of hours. So better stick with 0.9.5.3.

Note: make sure you have php5-dev installed otherwise compiling will most likely fail

#cd /usr/local/install
#phpize
#./configure –enable-eaccelerator=share –with-php-config=/usr/bin/php-config
#make
#make test
#make install

Then add (add the bottom) in /etc/php5/apache2/php.ini these lines
extension=”eaccelerator.so”
eaccelerator.shm_size=”1024″
;eaccelerator.cache_dir=”/var/cache/eaccelerator”
eaccelerator.enable=”1″
;eaccelerator.optimizer=”1″
eaccelerator.check_mtime=”1″
eaccelerator.debug=”0″
;eaccelerator.filter=”"
eaccelerator.shm_max=”0″
eaccelerator.shm_ttl=”0″
eaccelerator.shm_prune_period=”0″
eaccelerator.shm_only=”1″
eaccelerator.compress=”0″
eaccelerator.compress_level=”0″

Make sure you save and then restart apache with /etc/init.d/apache2 restart and that’s it.

Step 10 - configure RAID1 to send you alerts and make a test
Having a software RAID is good but it is even better to make sure it works properly and you’re alerted whenever a problem occurs.

Edit /etc/mdadm/mdadm.conf and add these 2 lines
MAILADDR your_email@wheretosendthealert.com
MAILFROM root@your-server.com

Now test that everything works properly by forcing one of the HDDs to be considered fauly
mdadm –manage –set-faulty /dev/md1 /dev/sda2

Now have a look at this HDD and you should see that it is not synced anymore
mdadm –detail /dev/md1
By this time you should have received an email with an alert that the HDDs are not mirrored anymore

Now that that you’ve teste that the RAID alert works fine just remove and then add the “defective” HDD:
mdadm /dev/md1 -r /dev/sda2
mdadm /dev/md1 -a /dev/sda2

Step 11 - finally, turn ON full power
Basically the server is by default configured to run in an economic mode, that is, it won’t run the processors at full speed. This may generally not be that bad if you suppose that the CPUs will run faster as their processing need increases but I chose not to risk :-)

So just edit /etc/default/cpufrequtils type in “performance” for “GOVERNOR” and then
#/etc/init.d/cpufrequtils restart

That’s all folks. If you need any help/have any questions just drop a message and I’ll be happy to help.

Advertisement:

Open Relay checker

Open Relay means basically that anyone on the net can use or abuse your mail server and send email from anyone to anybody.
While this is generally a bad thing  (as you can read more here) there are still mail servers that accept to relay emails either because they were not properly configured or because they where cracked.

The Open Relay checker tool is a tool useful for anyone that needs to test whether their email server accepts to send emails in an indiscriminate fashion.

How much “Unlimited” disk space/bandwidth is after all?

There is an increasing number of shared hosting providers that offer “unlimited” features, most usual ones being traffic(bandwidth), disk space and domains or websites. Well, this “unlimited” should be taken with a pinch of salt as some of you may expect.

I did a small analysis of two leading hosting providers:
HostGator and BlueHost.
Both providers are good, have excellent support and they offer a good value for money as I figured out myself working with them since over a year.

Anyway, one looking at their offer could wonder why should they buy more expensive dedicated hosting when they could for as little as $7-$8/month get unlimited bandwidth, disk space and so on?

The reason is that this “unlimited” is mostly a marketing catch-word and a person with some Linux/Unix administration experience, reading their Terms could more or less easily spot the tricky spot.
Shared hosting is basically offered on a powerful server usually with 2+ processors, 8+GB of RAM and some TB of disk space. The company then tries to squeeze in as many user accounts as possible. The better the hosting provider’s services are the less loaded the server is and the less troublesome for you is to run a more-than-basic site.

To understand this better you have to keep in mind what are the biggest bottlenecks for a server:

  • processing power
  • HDD space

While a server could serve thousands of plain HTML web pages per second an unoptimized database query run from by you PHP scripts could load the server significantly. Multiply a bad query with some 1000 of visitors/hour and the server will start to get slow…or VERY slow.
Sort of a similar thing can happen with disk space: while hosting 10 x 1GB files should not be pose problems for a server, 10 million x 1KB files will not only use A LOT more phisical disk space but it will also consume a lot of CPU when it comes to read/write/delete those files.
I tried this myself and only creating 1 million directories on an ext3 file system took a powerful dual server about 1 day!

Considering all of this you can imagine that the average shared hosting provider will try to indirectly limit those “unlimited” features. Otherwise they’ll lose money and that’s not very…capitalistic :-)

Now let’s get back to my 2 examples:
1) BlueHost chooses not to give numbers; they say they offer “unlimited” such and such features but they also explain that their service is intended for small business:
BlueHost.Com’s service is designed to meet the typical needs of small business and home business website Subscribers in the United States. It is NOT intended to support the sustained demand of large enterprises, internationally based businesses, or non-typical applications better suited to a dedicated server.
I find BlueHost’s Terms a bit unfair as one could not know when exactly they’ve reach the limit and their site will be suspended

2) HostGator is more clear in explaining what they consider a threshold that could bring penalties:
The use of more than 250,000 inodes on any shared account may potentially result in a warning first, and if no action is taken future suspension. Accounts found to be exceeding the 50,000 inode limit will automatically be removed from our backup system to avoid over-usage. Every file (a webpage, image file, email, etc) on your account uses up 1 inode.
So as long as you keep it under about 50,000 files you’ll have all their features ON; if you pass that threshold you could start getting into different troubles.

As I explained above the “unlimited” has its limits whether they’re clearly stated or not.
Make sure you take the time to read and understand provider’s Terms and don’t expect to be able to run CPU expensive/high traffic sites without running into problems.
Better get a dedicated server once you start reaching your shared hosting provider’s limits.

How to check your computer’s open ports

Many times in the past I needed to know whether some of my computer/router’s ports where open or not in order to be able to play an online game or to use eMule.

Let’s take eMule for example: in order for your computer to get a high id (and be able to make fast downloads) you need to have some TCP/UDP ports open.
As it took me time to make changes to my router I was frustrated to discover that I had to wait then more for eMule to connect to servers and then check whether I get a high or low ID.
Not to mention about the fact that sometimes the servers wrongly indicated that the ports were closed while they were open!

Anyway, after looking on the net and not finding a free online tool that could do something like this I decided to create my own port scanner:
http://tools.geek-tools.org/en/port-scanner/

Why is this good?
First of all because is mine :-)

Then because it works, it automatically detects your IP, you can choose to check any of your computer’s ports or make a 0-1024 TCP/UDP port scan so you make sure there aren’t  any open ports that could lead to vulnerabilities.

Automated directory submission: cons and cons

On this vast Internet there is an interesting niche that brings some good money with very little effort to those who offer the “service” but little positive outcome to the customers.
The nice that I am talking about can be called “promote your site massively for little money” and has among others the sub-niche “massive web directory submission for awkwardly little money”.

Today’s case study is: 5000.ro
As with these automated submission sites it comes at an excellent price: 6Euro for 1100+ sites.

To spare you the time, effort, money and frustration I’ll describe below my experience:
1) even if you pay only 6Euro for a submission you’ll have to validate for the next 2-4 weeks many of the automatically sent applications. Many of these directories you apply to send an email which you have to read and click on the validation link. Time consuming.
2) I fortunately used a dedicated email for the job. If you made the mistake to use your regular email address then expect 1000+ emails upon they started submitting your site and then a regular SPAM emails for the next of your email’s life!
3) After you spent time reading and validating emails while deleting tens of SPAM emails daily you’ll start waiting for results to come…and wait…and wait…and wait.
I believe I waited for half a year and the only clear outcome was that my site started getting less traffic from Google after a couple of weeks and didn’t recover for a couple of months!
4) even after 6months+ if I check the backlinks to my site from Google I don’t see anything except the 2 sites I made a link exchange with.
If you don’t believe me just visit their site and check for past client’s backlinks

The bottom line is: save your time and money.
If you still want to promote your site try an offer like this one: in the worst case scenario it will bring you nothing compared with a lot of SPAM and less traffic as in 5000.ro’s offer :-)

Buy cheap hits - goldmine or scam?

Have you ever come across sites like BuyHitsCheap.com that sells you “10,000 Guaranteed Unique Visitors. Real traffic, real people, potential customers.” ?
Have you ever wondered whether this is too good to be true?

Well…it is :-)
Most people would think that paying some $10-$20 for this isn’t such a big sum and would not hurt to give it a try…after all is just 10(or 20) bucks!

That’s the whole catch: pay a small amount of money and greedily wait for those 10K visitors to stumble upon your site and even click on your ads.
If that was true could even bring you money, just think of this:
you invest $10 for 10K “unique visitors”. If your have a site with valuable content you could have say an average CPC of $0.1 and with an average CTR of 2% you could have about 200 uniques clicking on your ads thus generating you $20.
So you may think like this “I invest $10 and earn $20, about 100% profit…note bad, just found myself an excellent money machine!”

Well if you haven’t yet “invested” the money into this then you may not want to do it anymore.

Here’s why:
- there is no such thing as something for nothing. If those sites were indeed that good they wouldn’t advertise their services on Adsense and other ad networks, would they?
I mean, why spend money advertising your services when you could just open yourself a 10K uniques campaign and even if only one in 1K visitors will buy your services you’ll still have a $90-100 profit
- some “buy cheap hits” sites won’t even send you more than 10K automated hits from the same computer but this will be good compared to
- have your site placed in an “automated traffic exchange” scheme that will generate seemingly real traffic but this won’t bring you much hits (if any). Instead they’ll much likely bring you a ban from Adsense and similar ad networks as they’ll sense the traffic is not genuine.

There are some other even nastier schemes that will eventually bring you some useless traffic but I believe you got the idea: stay away from this kind of schemes. At very best they’ll just cost you $10 and bring you nothing :)


Server Uptime Info

Few people are aware that one can easily figure out the server uptime without needing more than regular HTTP/80 port access to it.

For those who are eager to read here’s a RFC on how this is supposed to work.
For those more eager use a tool than to figure out how it works in detail check out this online server uptime check tool.
The online tool is based on Nmap which is a freely available tool network scanning and security audits.

There are number of uses for this tool one of them is to figure out the uptime for a hosting provider. Basically if the feature isn’t turned off on their server you can check the uptime for different hosting companies. Generally a longer uptime is better as it would mean less downtime.

If you own a site and you’d like to turn off this feature here’s a quick Linux HOWTO on how to do it:

edit /etc/sysctl.conf

and add this line
net.ipv4.tcp_timestamps = 0

and then run the command
#systcl -p

to make the kernel change active.


Adult sites ad networks

If you own a site and you’re looking for ways to make money then Google Adsense is generally the best option.
Anyway, if your site happens to be an erotica/adult oriented site then Adsense is out the question as there are most of the other advertising networks.

Looking on the Internet for good adult advertising networks I stubled upon 2 which I would recommend you to try:

Etology: a pretty straight forward way to start publishing adult ads on your site.
The system isn’t too complicated and that makes it easy to use but in the same time it lacks useful features like setting a minimum CPC and/or setting a default ad to show when they have nothing available.
My average CPC on this network is about 2 cents which for the adult industry is pretty good.

Madison Avenue: a complex (some may consider it even too complex :-) system with lots of features and options. Basically its a dream system where you can configure most of the things you could imagine.
There are only 2 issues I have with it:
1) it requires patience to configure a decent campaign
2) for me the CPC I got from them was about 1 cent so this may not be that good unless you take the time to configure everything properly and squeeze out more.

I would advise you to give it a try with both systems and figure out for yourself which one suits you best.
If you have good traffic coming from North America/Japan/Korea/Western Europe then you’ll start making some good money soon! :-)

Virused domains by country

In collaboration with Elbia Scan we started scanning for all the virused sites by country.
So far we collected data just for 6 TLDs but more will be available soon. Once we gather more data we’ll try to create statistics to see which are the most dangerous TLDs and also which are the most vulnerable IPs.

Here are the links to the first 6 TLDs analyzed so far:

  1. Romanian (.ro) virused domains
  2. Czech (.cz) virused domains
  3. Slovakia (.sk) virused domains
  4. Commercial (.com) virused domains
  5. German (.de) virused domains
  6. Russian (.ru) virused domains