Recently one of our readers asked us to post an article about the benefits of Windows driven VPS hosting. Since we’ve been looking into provision of this service as well, there was some interesting information gathered and we would like to share this with you. We thought that showing both upsides and downsides of this type of hosting is more fair. Sure, this article is posted from our point of view, so it reviews peculiarities of Windows hosting in terms of use of VPS-hosting specifically as a platform.

How is Windows run on VPS servers?

Running Windows on VPS usually implies the use of two types of virtualization technologies – XEN – the technology, generally applied in creation of fully virtualized VPS platforms and Hyper-V – special virtualization system, developed by Microsoft to run Windows VPS server only. XEN is known to many of our readers – this virtualization technology allows to create and manage virtual servers keeping them completely isolated from each other, while using same hardware resources by different system kernels – each fully virtualized environment keeps the system with its kernel that addresses to the layer of virtualized hardware.

As for Hyper-V – this technology needs some description. In fact, it is a hypervisor that works on the basis of partitioning with at least one parent partition that has access to hardware and can create child partitions that host guest OSs within virtual environment – just like XEN and its interfaces. But if you take a closer look, you can see that Hyper-V is a version of core installation of Windows Server 2008 with virtualization support – which makes it look much like OpenVZ that utilizes modified Linux kernel. Thus, it is fair to call it a symbiosis of two technologies.

What is so special about Windows VPS Hosting?

VPS as a platform to run Windows on has become a real discovery for many users who needed Windows-based hosting but could hardly afford a dedicated server while experiencing issues with shared hosting, run under this OS. Indeed, using Windows as a server OS and namely for hosting needs is quite troublesome – initial resource usage is higher than that under Linux, disk usage policy and work of the disk subsystem is more complex, and full functional access can be achieved only by establishment of remote desktop connection (RDP) or a control panel. The latter aspect caused the biggest inconvenience, since shared hosting users are not usually provided with RDP access while the control panel functionality is limited so, that the lack of privileges makes the use of many tools impossible.

Use of VPS hosting has solved many of those problems – VPS carriers with large amounts of RAM allowed to delegate resources that are enough for the system to operate with additional software and cover the loads created by resource intensive programs (e.g. Microsoft Exchange server); RAID hard disk arrays + virtual disk partitioning allowed to minimize the I/O loads and increase the read/writing speed. Finally, a VPS allowed to create isolated environment, which enabled the users to manage their servers under “Administrator” privileges.

What are the downsides to keep in mind?

Besides the obvious upsides of Windows-based VPS hosting, there are, however, downsides that should be considered while making the choice. First of all the price – when it comes to evaluation of Windows hosting in general and its hosting on a standalone server in particular, all licenses, starting from the OS itself and all software, seem to cost an arm and a leg. The server is not going to be cheap as well – as mentioned above, Windows requires more resources, so the least powerful virtual server you will need is a XEN-based VPS – which are more expensive than those under OpenVZ – with 1 GB of RAM at minimum (you can find plans with 512 MB of RAM offered, but there won’t be much resources left to operate with). A Hyper-V based server with same resource rates will grant you more stable hosting, but it is going to cost more.

Mentioning this downside we don’t mean that the pricing difference is going to reflect in tens of dollars compared to Linux VPS hosting, but this may make some users think, whether they need it – at least those, who are not sure, if they are going to prefer ASP and .NET to PHP. If you are sure about the use of Microsoft software, here are some more tips on what you should be careful about.

Accessibility. You may be happy about having RDP and control panel access to your server – both provide clear intuitive interfaces, the former actually makes you feel like sitting at your home desktop, but are those means actually enough? If your server gets too overloaded, you may not be able to access it to reboot it, so you will need to send a support request which definitely causes a time loss. What you may make sure about before choosing your Windows-driven VPS host, is if they offer remote reboots from the host’s client panel – this may save your time.

Monitoring. Unlike Linux-based VPS servers, those running Windows need more severe monitoring, which may result into either bigger expenses – so you will need to hire an administrator or involve your host’s support team – or stresses, caused by constant inspection of your server’s state and investigation of issues it has or may have in future.

There are even more aspects, related to OS Windows itself to be kept in mind – security, as Windows is quite vulnerable; software-related; usability-related and others. So it is fair to get back to the question, whether you really need it. Many webmasters who are asked to create a site to work on a Windows server usually ask this question. And when they receive a solid “yes” they want the project to be described as clearly as possible. Thus, if you imagine your project well and see the full picture of how it is going to be run and why exactly it needs Windows, you are on the right way, and we hope, that your experience in this is going to be positive.

We are back to the topic of cloud hosting. The latest article we published on it was dedicated to two cloud hosting platforms we compared – a standalone server with CloudLinux OS installed and a cloud, based on several VPS-nodes. Since we analyzed both platforms from the point of view of shared hosting, it is now time to take a closer look at what cloud-based shared hosting actually is like.

The main concern there is about resource delegation – those who use shared hosting, know that providers of shared hosting services usually limit server resource usage softly – they announce the limitations in their Acceptable Usage Policy or any other document, that regulates service operation, which means that anytime the user abuses the server by exceeding the limits and overloading the server, they can suspend the account.

Does this always work? Unfortunately, no. There are many cases when the server goes down simply due to a rapid load spike that can hardly be maintained immediately, so the administrators don’t manage to do anything to reduce the load or lock out the abusive account. This may cause fifteen-minute downtime at minimum – fifteen minutes just for the server to reboot. It can however take more time to get the server back online – procedures of File System Check and rebuild of the RAID array may last for several hours. And they cannot be skipped. Such unbearable downtime may ruin the hosts’ and their clients’ businesses overnight, therefore less hardware-dependent solutions are needed. But what if we take such hardware-dependant platform as a single server working under CloudLinux and compare it to a simple shared hosting server?

The first thing that you are going to notice is the user limitation option – this unique ability of the CloudLinux OS minimizes the risks of server overload since each account is isolated owing to the kernel level technology in use. If you used to suffer from abusive neighbor accounts on your former shared server – you are going to like it. But what if your online project is active enough? You might have caused some temporary hard-to-mention overloads on your good old shared server – everyone does – moving to the server with CloudLinux you need to watch it. An isolated account may simply hang after reaching the resource usage limit, so you will have your “local” downtime with the entire server staying online, and you may not be able to have it fixed unless you contact the support team.

This brings up the first tip: consult the sales department of your future cloud host to find out what the user account limitations are. If they provide relative percentage rate – request the server hardware specifications – although those are not going to tell much, you can estimate the amount of available RAM you can use risk-free.

Another peculiarity to get acquainted with while moving to CloudLinux-based shared server, is the limitation of resources for certain applications. Depending on resource intensity of the software in use, administrators may set limits on CPU/RAM usage for the most demanding apps. Sure, no limits are going to be set by default, and the hosts do not actually interfere into work of their clients’ accounts, but this option may be applied if needed. For instance you can request the limitation of your forum, blog or gallery if you expect a great number of visitors each time new content is uploaded – this will save your web, email and other services under your account from hanging. Or you can ask to set a permanent limit for some of your scripts with lower priority to keep the most important application constantly up.

For this option to do you good you need to keep another tip in mind: check your visitors’ statistics and average resource usage rates, consult your technical support team if needed to figure out your resource limitation strategy.

The last but not least is multiple domain hosting. After introduction, the idea of shared hosting with multiple addon domains support influenced the significant increase in sales – people liked the possibility of purchasing one account and host numerous websites there. Considering the way CloudLinux works with resource distribution, it may be reasonable to try separating you websites to keep them under different accounts.

Conclusion

Shared hosting on CloudLinux may look strictly organized, but it does cut some dependencies – you don’t have to worry about other clients’ resource usage and even about particular software that may sometimes be intensive for a shared server. As for performance – it can hardly be called limited – first of all the servers for CloudLinux are rather powerful, so the so called “fair share” of resources assigned to your account will be enough for your websites. Another thing to keep in mind is rational account usage – if you feel that some of your scripts becomes too intensive, it may be worthwhile limiting it in resource usage on time or getting it transferred to a different account on the same server. Thus, if you follow the rules and common sense you are going to have your sites hosted without any problems.

Not that long ago we published an article introducing you into cloud hosting. This relatively new, yet powerful and promising technology becomes more widely used in hosting of small and medium online projects day by day. It means that this kind of hosting services gains users trust and requires less technical skills than one might think. Additionally, it makes prospect customers expect further price reduction, which in turn may increase the sales of this product.

If you want to make sure that all your plans regarding your hosting ground get realized, it is vital to do some pre-sales work targeted at finding out as much technical information as possible. Many things about the technology in use are usually available in extended feature list of packages offered by a hosting provider, at their corporate Wiki or FAQ documents. Today we are going to talk about the primary parameter to look after while choosing your cloud host – platform.

Platform for Cloud Shared Hosting

Mentioning small to medium websites was done on purpose, since the owners of such projects form the main layer of shared hosting customers. Being well optimized many websites, even relatively resource intensive ones, can exist on a shared server not requiring any upgrades and without causing any overloads. But what if the account on the server caused minor loads or simply produced load spikes during unexpected visitor rate increase? Virtually it might stay on the same server, if the latter had more resources… or if the account had a fixed limit on each server resource provided. Of course, this would cause slowdowns and even short downtime for the specific account holder, but considering the fact no other server users were affected, no one would ever suspend such an account.

In fact, we really get a shared hosting platform capable of hosting the same or a larger number of users than on a simple shared server and enhanced with dedicated resource distribution – this sounds interesting. Which technology do they use for such solutions? Since it’s a shared hosting platform, you will hardly find any clustered solutions with a lot of physical servers involved. There comes a question, how to create a simpler yet reliable cloud hosting platform, using one physical machine. Composing this review we decided to compare two solutions of such kind.

Standalone Server with CloudLinux OS

If you give a quick look at what the CloudLinux OS is, you will find out that it’s just a CentOS-based build with special patches. The way those patches work, however, truly confirm the “Cloud” prefix in the name of this OS – namely they utilize a kernel level technology enabling per user resource limitation. As a result, no account can use more than the fixed amount of resources assigned. Reminds of the way OpenVZ VPS servers work, doesn’t it? Yes it does, owing to CloudLinux OS all server users are isolated from each other so anytime some of them go down, all the other user accounts stay stable.

Apart from those user-level limitations CloudLinux supports resource-per-application limits adjustment. Thus, if you have several projects hosted under one account, you can distribute the limits, depending on websites priority. Say you’ve got a serious business website that has to be constantly online and you have a corporate blog that may cause load spikes during promos you run or whatsoever. Well, you limit your blog, so next time it’s overloaded it simply goes down on its own letting your business website keep working with no interruptions.

As far as you can see, CloudLinux OS is something that is worth being given a try as a cloud hosting platform. It however has a sole disadvantage that may stop you – your account exists within single environment, so any OS corruption may cause serious downtime or even data loss.

VPS Cloud Hosting

Unlike CloudLinux, VPS Cloud platforms do not provide single environment. Well, the environment you work with is actually single, but in fact it is supported but several VPS servers (VPS nodes). The number of nodes forms the amount of resources available for you. The more nodes you have, the more powerful your platform is… and the more redundant. If your CloudLinux environment goes down, you’ll experience downtime unless the issue is fixed. VPS Cloud never goes down off hand – each node takes care of your environment, restoring all services with minimal downtime, then the nodes that have failed recover and your system gets completely stabilized.

Building cloud on VPS first of all means providing utmost scalability. You can adjust the number of available nodes anytime you need more power, or vice versa – less expenses. Taking to account the initial cost of VPS-based cloud hosting – such scalability can really be considered vital. You may wonder why VPS cloud hosting is so expensive compared to CloudLinux hosting solutions – the reason is hardware. VPS cloud hosting is based on XEN hypervisor technology that implements usage of a more powerful carrier server. Taking into account the fact that each carrier should be able to bear over 20 VPS nodes at minimum – for optimal ROI, the physical machine requires a great shared storage – this involves usage of RAID array solutions with fast (usually SCSI) disks.

What you get for all this – is a completely isolated environment with a way more powerful resource base initially ready for higher loads. Even the lack of root access doesn’t sound like a disadvantage – you can anyway have your environment adjusted, at least with the help of the technical support team of your host.

The Bottomline

Although two platforms look different, you can use both of them for cloud hosting. Can one really call CloudLinux a platform for cloud hosting? Yes, if you need to provide affordable shared hosting with enhanced stability. What is VPS cloud then? It is a solution for more resource intensive projects, but if you are a reseller – you get the same platform as those who choose CloudLinux. Definitely, you are backed by a load of nodes that make your service rock solid but from the point of view of the end user both platforms will look like inexpensive hosting grounds with improved stability and reliability. Thus, if you are looking for a reliable cloud hosting solution – you should check out the platform the service is based on. CloudLinux is going to be cheaper; VPS based cloud hosting may be more expensive, though powerful and redundant, but you can be sure – both services are based on utmost high performance hardware that is monitored and thus secured from failures.

If referring to our personal point of view – Cloud VPS is a more redundant and reliable solution. Taking into account the fact that we at SiteValley.com are looking into starting offering cloud hosting, too – let us announce that having analyzed the given niche we are on our way setting up the first VPS node.

Have you ever purchased a hosting solution that in the end appeared to be either not sufficient for your needs or too powerful than actually required? Well, if you did, you are most likely to have experienced inconveniences and service slowdowns in the former case, while the in latter case you realized you are paying too much extra. Thanks to the technology called cloud hosting, it is right now easier to get a hosting solution that would provide the exact amount of resources needed at that particular point.

What Is Cloud Hosting?

Cloud hosting, also known as cloud computing is a relatively new technology that involves usage of a multi-server platform, which allows create cyberspace and process data there without snapping to particular hardware. In another words, it doesn’t matter which particular physical server your web project is hosted on, since there are many servers working in a “cloud”. You wonder how it is created? In fact, a “cloud” is a virtualized environment based on several physical servers. The virtualization technology used for cloud hosting is similar to the one, used on VPS hosting; cloud, however, works contrariwise – it requires several physical machines to be united under the same virtualization infrastructure.

Cloud Hosting Servers Scheme

VPS Hosting Servers Scheme

Virtually, the idea of cloud hosting is a realized concept of full service clustering. Use of cluster systems in hosting is very popular – some hosting providers put their nameservers on a DNS cluster to provide better availability of the DNS servers, some website owners use storage cluster systems to improve data fail safety. You can even create a separate database cluster if needed. The idea is that any solution of those significantly improves fail safety, stability and accessibility of the vital services. Cloud hosting providers consider all services vital; therefore any single service, process or piece of data is not bound to hardware. This grants cloud hosting users something, that VPS owners are familiar with – no hardware issues. Unlike a VPS, however, cloud hosts are safer on that point. A VPS carrier may fail and disable a virtual machine – cloud hosting shows better fault tolerance – the more physical machines support the cloud, the more servers can be temporarily unavailable at a time.

Use of Cloud Hosting

If we analyze how Google or Facebook work, it becomes obvious that cloud hosting is the best solution for websites with huge amount of daily traffic that operate large amounts of data by lots of simultaneous visitors. Well, even if your online project is not a great email exchange service or a social network you might find cloud hosting useful, e.g. if you run intranet collaboration software or image gallery, or a complex website with corporate blogs, forums and other applications.

Referring to the first paragraph of the article it is vital to note that cloud hosting may become a perfect platform for any site, regardless of whether it is hosted on a shared or a dedicated server. This is first of all determined by utmost scalability of cloud hosting – you get as many resources to support your platform as you need. Thus, moving from a shared hosting package to a cloud hosting plan with little resources, you should not expect big difference in pricing, while the performance and stability are going to grow. The same refers to dedicated servers – choose the optimal specification and enjoy your new hosting platform.

You may wonder what operating systems are supported by cloud hosting platforms, since some applications and languages can operate under a certain OS only. Right now both Linux and Windows-based cloud hosting services are available. Additionally, both types of service are competitively represented on the market so there’s no problem with finding the best platform for you. You may remember us revealing some peculiarities of Microsoft Windows Azure Cloud Platform; right now there many more other platforms available and we are going to release a review of some Linux-based platform shortly, since Linux still remains a better option for webhosting solutions.

Advantages and Disadvantages

Even if you are going to use cloud hosting as a simple shared hosting solution for a single website of yours, even if it is provided by cloud hosting resellers, you are going to get a significantly more reliable solution. This is owing to a number of advantages of cloud hosting services, namely:

Scalability – we named this advantage before, and it’s hard not to highlight it once again. Apart from the ability to host a wide range of web projects from a personal web page to a serious online business website, scalability of cloud hosting allows perform upgrades painlessly without data transfers, downtime and DNS propagations.

No hardware issues – this one has been also mentioned. Improved fail safety is one of the main benefits of cloud computing. Regardless of the hardware issue, taking place on one of the servers within the platform, the availability of your website is hardly to suffer.

Management – being a complex system, cloud hosting platforms are maintained by professionals so you don’t need to think of any hardware or platform issues. Taking into account the benefits mentioned above cloud users are not likely to need contacting technical support regarding some major issue.

As any technology, cloud hosting also has its disadvantages. Usually they are connected with decreased security level and indirect data access – both due to virtualized file system. Some users also address to the lack of administrative privileges, but this concerns resold cloud hosting mostly.

Summary

If your web project requires an ultimate solution that would set you free from any concerns – cloud hosting is definitely something to consider. Having gained much experience in VPS hosting, we at SiteValley.com look into provision of cloud hosting as well. Thus, you should expect some more articles from us in this niche, including our own conception of services we are going to launch.

Starting reseller business five years ago or earlier the business owner had only two options: signup for a shared hosting account with reselling feature or order a dedicated server. The initial option was a fine solution for starters since it was cheap, but as soon as the resellers’ business developed they needed a step-up. The problem however was that a dedicated server was still not affordable for them, while the resource usage of both resellers’ and their resold accounts became completely unbearable for a standard shared servers. This led to continuous suspensions of accounts and constant quarrels, some of which were even brought to social media. Finally, either the hosting company met halfway with the reseller offering some discount on upgrade or the reseller quit the company to signup for another shared hosting plan with another host. A lot has changed since VPS hosting has been introduced, so nowadays hosting reselling is not that tough, expensive and risky business.

What is a VPS and How It Works?

A VPS is a Virtual Private Server – an isolated virtualized environment with its own OS (Operating System) and emulated hardware, hosted on a physical server a.k.a. carrier. Those virtual servers are setup within a carrier by means of virtualization technology, which allows to run several isolated boxes, letting them operate independently from each other and provide the performance rates, similar to the ones of a dedicated server. Though there are several virtualization technologies, the common idea implied in all of them is to create conditions, on which several operating systems could function within single hardware set – one physical machine. There are many backgrounds for such an idea – energy use reduction, hardware use optimization, setup and maintenance cost cutting, etc. The main thing is that this idea works well for many business solutions including hosting reselling – let’s take a closer look, at which aspects specifically.

Reasonable Price

An upgrade from a shared reseller plan to a VPS or initial start of reseller business on a virtual server is quite an affordable solution, since it sets the business owners free from enormous expenses of hosting. Resellers can then invest their money into SEO and social media marketing to have their business promoted on the World Wide Web. Apart from just being a good fund saving solution, a VPS stimulates the reseller to fair pricing formation. This is one of the key points in this business, since, on the one hand, there are too many competitors, while on the other hand – spending much money on expensive hosting may force the reseller to set high prices, too high prices for the provided services.

Smooth Upgrades

Taking into account the further growth of their business, resellers may need the further upgrade of their VPS. The virtualization technology allows performing such upgrades staying on the same server, by just increasing the resource usage quotas – to some extent of course. Thus, since VPS hosting providers usually offer 3-5 different plans, a reseller may have his or her virtual machine upgraded to a couple of levels higher. Such upgrades are very convenient as long as they cause no downtime, which is critical for resellers and their clients.

Master Reselling

The feature of master reseller plans provision – plans which allow reseller’s customers to resell hosting, too – was usually available for the dedicated server users only. Modern technology allows to offer such plans on a VPS. Of course, this solution requires a more powerful VPS, but the fact is that resellers can now provide reseller hosting on their own with less expenses involved.

Virtual Hardware Makes No Real Issues

One of the main benefits of VPS hosting is the absence of hardware issues. This point is the most essential for resellers. If the hosting reselling business owner is using a shared server and this server crashes, a reseller depends on his/her hosting provider, who in turn depends on datacenter staff. As far as such issues may take long to get resolved, a terrific waste of time may cause resellers to lose their customers. If it is a dedicated server, the issue may take less time but more money (e.g. if there’s no SLA guarantee, if the server is purchased from the datacenter directly, etc.). Being hosted on a VPS a reseller is set free from any possible hardware issues, except for those which may happen to the carrier. In this case, however a Virtual Machine can be cloned and moved and then started on the backup server.

Server Management

Performance is not the only feature of the VPS, which makes it look similar to a dedicated server. Another feature is management. Administration of the server does take much time. If there are a couple of dozens accounts hosted there, administration and management of such a server may become a nightmare. Those who were using a shared reseller hosting account can tell how exhaustive it can be – to just manage clients’ accounts and billing. Imagine what it would be if they also had to administrate the server itself. Well, you may not imagine this – virtual servers are administrated by the host’s technical staff.

Another positive thing about VPS management is the client side. Let us refer to shared reseller hosting users again – I believe all of them can confirm they had a situation, when they requested assistance for their customers, even on their behalf and received a deny. This may sound fair enough from the host’s side in terms of support provision policies or taking the fact the server is shared and is used by many other people and so on, but the problem remains and the reseller’s client stays unsatisfied. Use of a VPS helps to resolve such issues quite well. First of all, a VPS is one of the top products at many companies; therefore its support is privileged. Another point is server isolation – it allows the host’s support team to modify initial server configuration and apply non-standard solutions. As a result – the reseller can take care of many other things important for his or her business.

Customizability

Another advantage of a Virtual Private Server for a reseller is the ability to customize it. The isolatability of the VPS and the presence of root access to it allow the reseller to tune-up and reconfigure different services, what helps minimizing load and cut server resource usage. Considering the constant lack of disk space and taking into account active file transfer and mail exchange, such option may do much good to both reseller and all his or her clients, since a couple of resource intensive accounts may make other account owners and the entire VPS experience services slowdown and even overload. For instance, a reseller may limit the number of FTP and SMTP/IMAP connections to the server to lessen the number of request and the load on those services, in turn. Furthermore, the administrator of the VPS may simply disable IMAP protocol on the server, so that all the users could use only POP3, which implies the download of all messages to the local computer of the user. Such limitation helps saving server space and gives the reseller a ticket to overselling.

Overselling

Overselling is actually not the best strategy, but many hosting business owners apply it while creating their hosting packages and setting prices for provided services. In a nut-shell, this strategy means offering and selling more, than one really has. This mostly concerns disk space and bandwidth quotas, but hosting sellers and resellers usually oversell on more services – number of virtual hosts (addon and subdomains), email and FTP-accounts number, size and number of the databases allowed under a single account. Though overselling strategy does not always bring many problems, the risk of slowdowns and overloads may grow with its use. Thus, if the hosting business owner considers this solution important for his/her business, overselling may be applied. The reseller however should make sure the server in use is capable of carrying out such loads.

Modularity

The problem of load and its reduction has always been in question for all server administrators. Even for those who use their server resources rationally. The solution to this problem was found since the load balancing technology had been invented. This technology requires the use of several servers – each of them is processing requests addressed to a service or a set of services specifically configured on this standalone machine. Load balancing may look an expensive solution, but its price is justified by the positive effect it makes. Additionally, talking about separate machine for service processing, we mean virtual machines, too. Those are significantly cheaper, as far as you know, so let us see how resellers can have load balancing applied in their business.

Imagine you have a top VPS. To make the story more interesting, imagine that you oversell and already have about 50 clients. You start experiencing load issues because some host over 20 domains; some use your service as a small email server and some organize file sharing over FTP. Such load is bearable yet, but you realize that the further growth of your business may require an upgrade – this time to a dedicated server. You are definitely not interested in wasting time on transfers, DNS-rearrangement and other related problems. Thus, you get a couple of simplest cheaper virtual servers, one of which is let’s say set to process MySQL requests, while another deals with mail exchange. Even such move is going to seriously lessen the load on your VPS.

Summary

The development of the reseller’s business may later require an upgrade to a more serious solution – to a dedicated server. A finely tuned VPS however can let a reseller host his resold accounts for a long time with no issues and for a significantly affordable price. If you are not sure about what hosting plan to choose to start your reseller campaign – order a VPS and start seizing the market of web hosting.

Scripts Resource Intensity

Nowadays having a personal website is as common as having a cell phone. Some people use cell phones for just giving and receiving calls and messages and some cannot simply imagine their mobile telephony device without a camera, an MP3-player and so on. It really is likewise about the websites: some website owners use their hosting accounts to bring up pages on the web and use email at their own domain. The others want their websites to be an ultimate public project, which would support many options starting from integration with social networks and ending with, but not limited to provision of downloadable content.

The way the sites are built is a question of demand, of course. But this question raises another one in turn – a question of approach. Many website owners start with shared hosting but then face with the problem of resource overuse, as their projects keep developing. And that’s where the question of approach becomes essential – those who take such issues seriously usually decide to upgrade in order to avoid temporary suspension due to excessive resource usage.

Of course, the website owner may not always realize the influence of his/her particular account on the entire server. On the other hand, server administrators are always glad to assist with the resolution of the problem. Those suspensions are mostly a preventive measure, which doesn’t let the entire server go down. For example, our team is always open for a dialogue and we co-operate with our Customers to find out the way to get the issue resolved. However, if it is definitely not about some particular script or module which is enough to be disabled to let the account keep working on the shared server – an upgrade is the next step.

VPS as a Hosting Ground for Resource Intensive Websites

Several years earlier the word “upgrade” would definitely mean a setup of a dedicated server, which would be tens times more expensive, than a shared hosting plan you used to have. Modern technology, however, has introduced a more liberal solution – Virtual Private Server (VPS). Those virtual servers are containers, created by means of virtualization software on physical servers. Current platforms allow a VPS web hosting user to obtain almost the same level of performance as the server-carrier provides, which means that a user can get a dedicated server, though a virtualized one, for a significantly cheaper price.

A virtual server has the following key advantages, which are to be considered during the upgrade:

1. Isolation – none of several virtual servers on the carrier influences each other. No limitations are set.
2. Full root environment – the user is provided with the maximal administration privileges.
3. No hardware dependencies – server does not require file system checks and RAID array rebuilds after reboot.

As you can see, a VPS resolves the main concern of a shared hosting account user – mutual user dependency. The thing is that shared servers have such strict limitations primarily due to the problem of even resource distribution, i.e. those limits create fair equal conditions, which let any user run his/her applications without abusing the other accounts. This equality is also guaranteed on a server software level – all major services, like Apache, MySQL and PHP are configured in a standard way and can be slightly tuned up only on a user level by means of local configuration files (e.g. .htaccess, php.ini, etc.).

VPS web hosting, however, provides each user with isolated environment. This means, that there is no one to share resources with, i.e. there are no limits and that any server software may be tuned up according to the peculiarities of the hosted script/application.

Optimizing your WordPress Blog Performance

There are many scripts which require certain tune up. The more tasks a script is to carry our, the finer the tune up should be. As an example we take WordPress blogging tool, which is known for its resource usage peculiarities and a wide range of available plug-ins.

Surely, WordPress is one of the greatest scripts on the Web. This blog tool lets you create a nice and usable website by means of a user-friendly easy-to-comprehend interface. Its basic configuration is fine to be hosted on a shared hosting server and we offer a special plan Blog Hosting with WordPress preinstalled to those, who would like to start their own blog. Still, this platform is a rather resource intensive one, so depending on your goal and means of project realization, you may need to have your account upgraded.

Of course, it is essential to find out, what exactly makes WP resource intensive. Primarily it is the way, it works with the databases. Below you can find an example of a MySQL request created by one user, who is viewing one WordPress page:

3 Query       SELECT option_name, option_value FROM wp_options WHERE autoload = ‘yes’
3 Query       SELECT option_value FROM wp_options WHERE option_name = ‘rewrite_rules’ LIMIT 1
3 Query       SELECT SQL_CALC_FOUND_ROWS  wp_posts.* FROM wp_posts  WHERE 1=1  AND wp_posts.post_type = ‘post’ AND (wp_posts.post_status = ‘publish’)  ORDER BY wp_posts.post_date DESC LIMIT 0, 10
3 Query       SELECT FOUND_ROWS()
3 Query       SELECT t.*, tt.*, tr.object_id FROM wp_terms AS t INNER JOIN wp_term_taxonomy AS tt ON tt.term_id = t.term_id INNER JOIN wp_term_relationships AS tr ON tr.term_taxonomy_id = tt.term_taxonomy_id WHERE tt.taxonomy IN (‘category’, ‘post_tag’) AND tr.object_id IN (10) ORDER BY t.name ASC
3 Query       SELECT post_id, meta_key, meta_value FROM wp_postmeta WHERE post_id IN (10)
3 Query       SELECT * FROM wp_posts  WHERE (post_type = ‘page’ AND post_status = ‘publish’)     ORDER BY menu_order ASC
3 Query       SELECT option_value FROM wp_options WHERE option_name = ‘page_for_posts’ LIMIT 1
3 Query       SELECT * FROM wp_users WHERE ID = 1 LIMIT 1
3 Query       SELECT meta_key, meta_value FROM wp_usermeta WHERE user_id = 1
3 Query       SELECT * FROM wp_comments WHERE comment_post_ID = 10 AND comment_approved = ’1′ ORDER BY comment_date_gmt DESC
3 Query       SELECT * FROM wp_comments WHERE comment_post_ID = 10 AND comment_approved = ’1′ ORDER BY comment_date_gmt DESC
3 Query       SELECT * FROM wp_posts  WHERE (post_type = ‘page’ AND post_status = ‘publish’)     ORDER BY menu_order, post_title ASC
3 Query       SELECT t.*, tt.* FROM wp_terms AS t INNER JOIN wp_term_taxonomy AS tt ON t.term_id = tt.term_id WHERE tt.taxonomy IN (‘category’)  ORDER BY t.name ASC
3 Query       SELECT wp_comments.* FROM wp_comments JOIN wp_posts ON wp_posts.ID = wp_comments.comment_post_ID WHERE comment_approved = ’1′ AND post_status = ‘publish’ ORDER BY comment_date_gmt DESC LIMIT 15
3 Query       SELECT COUNT(comment_ID) FROM wp_comments WHERE comment_post_ID = 12 AND comment_parent = 0 AND comment_approved = ’1′ AND comment_date_gmt < ’2010-02-22 16:54:11′
3 Query       SELECT COUNT(comment_ID) FROM wp_comments WHERE comment_post_ID = 12 AND comment_parent = 0 AND comment_approved = ’1′ AND comment_date_gmt < ’2010-02-22 14:25:24′
3 Query       SELECT COUNT(comment_ID) FROM wp_comments WHERE comment_post_ID = 12 AND comment_parent = 0 AND comment_approved = ’1′ AND comment_date_gmt < ’2010-02-08 18:51:45′
3 Query       SELECT COUNT(comment_ID) FROM wp_comments WHERE comment_post_ID = 12 AND comment_parent = 0 AND comment_approved = ’1′ AND comment_date_gmt < ’2010-01-29 02:58:05′
3 Query       SELECT COUNT(comment_ID) FROM wp_comments WHERE comment_post_ID = 12 AND comment_parent = 0 AND comment_approved = ’1′ AND comment_date_gmt < ’2010-01-29 00:23:01′
3 Query       SELECT YEAR(min(post_date_gmt)) AS firstyear, YEAR(max(post_date_gmt)) AS lastyear FROM wp_posts WHERE post_date_gmt > 1970
3 Quit

MySQL queries of the *SELECT type are one of the most abusive as they go over all the databases to find the needed value. This results into aggressive disk subsystem and CPU usage. By the way, the log above was for a simple WordPress page, which had no add-ons or plug-ins, so you can imagine, what it is going to be, if WordPress is run full-throttle.

It is absolutely logical to search a way for optimization of such queries, as far as the fewer queries are produced, the more free CPU time is left for other server services. That’s what caching is used for. All you need is to edit your wp-config.php file by adding the following two lines there:

// Enable the WordPress Object Cache:
define(ENABLE_CACHE, true);

But this is minor SQL optimization. Another reason for WordPress issues is webserver overload, caused by simultaneous http requests. WP does not offer any other built-in caching tools and the one –WP Super Cache, which is available as a plug-in – is rather resource intensive, so it is allowed for use on VPS and dedicated servers only (AUP, p. 10.1). WP Super Cache is a static caching plug-in, which generates html files that are served directly by Apache webserver without processing comparatively heavy PHP scripts. This plug-in is one of the means to increase your WordPress blog performance significantly.

Still, this is not all what you can do for webserver optimization. Now you should edit your Apache configuration file (full path: /etc/httpd/conf/httpd.conf) as follows:

Note: Before you proceed with file editing, make sure you backed-up your configuration file.

# Timeout and Keepalive
Timeout 30
KeepAlive On
MaxKeepAliveRequests 100
KeepAliveTimeout 6

#Maximum Client Connections
&lt;IfModule prefork.c&gt;
StartServers       8
MinSpareServers    5
MaxSpareServers   20
ServerLimit      512
MaxClients       512
MaxRequestsPerChild  4000
&lt;/IfModule&gt;

Note: MaxRequestsPerChild parameter is a tricky one, so 4000 is not the fixed value, you may try a range from 1000 to 4000 to see, which configuration is better and leave it.
Note: #Maximum Client Connections may be not available in the configuration file, if “prefork” Apache module is not configured (if not, there’s a “worker mpm” module installed instead).

Once the file is edited, restart Apache with the following command: /etc/init.d/httpd restart.
Additionally you may consult more online guides on disabling those Apache modules, you are not using.
Such modules take CPU time and RAM to get loaded and then simply do nothing, while you can comment them out in the configuration file.

Another service to edit is PHP. Find a global php.ini file (should be located there: /usr/local/lib/php.ini) and copy it to the root folder of your WordPress. Open it for editing to set the following parameters:

;*Turn off for performance
register_globals = Off
register_long_arrays = Off
register_argc_argv = Off
magic_quotes_gpc = Off
magic_quotes_runtime = Off
magic_quotes_sybase = Off
;*Allow PHP to accept large data
post_max_size = 8M
file_uploads = On
upload_max_filesize = 8M

Conclusion

We really hope you find these tips useful. Those are not the only ones, of course, so you can find many other ways of WordPress optimization, which may come in handy. It is still advisable to consult your technical support team before making any changes to the essential services on your server and of course, do not forget to backup your initial configuration files.

Using web hosting services normally presumes storing content located on a remote server that is available online 24x7x365, which means that such a hosting server is powered and connected to the Internet constantly. A very important notion in web hosting technology is uptime. Wikipedia defines it in the following way: “Hosting uptime refers to the percentage of time the host is accessible via the internet“, therefore, if a customer is paying for hosting services, it is crucial for him to expect and receive the maximum uptime values. Although hosting companies do apply various monitoring techniques to track server’s state at any given moment of time, no technology is perfect and thus it is important to perform your own independent uptime monitoring, which may come in handy in emergency situations. So how do you measure uptime values correctly? Should you expect a 100% uptime? What web site monitoring tools are available to a regular user? These questions are not as simple as they appear at a first glance. Information in this article will be useful to those interested in understanding and monitoring uptime of an account or server.

Who should pay attention to web host uptime?

To begin with, let’s find out why we should be interested in having maximum uptime. Importance of this parameter greatly depends on the goals assigned to your hosting account or server. Users with personal blogs, forums or regular web pages do not worry as much as people doing e-commerce, managing blog for marketing purposes, providing advertising services with your content, using hosting services as communication system for business, sending customers to affiliate programs, etc., as for the latter type of hosting customers the server downtime impacts their income either directly or indirectly. In case your profit, reputation or status depends on the availability of your content, it is a very good idea to perform uptime tracking.

Nature of server uptime

As it has been stated above, server uptime is its availability on the Internet, but the server is not a solid entity that simply appears on the internet as an object. There are always various services running on a web host. They ensure presence and correct presentation of data on the network. Major services handled by *nix (here *nix goes out for UNIX, Linux and FreeBSD driven servers) servers today are:

• Apache server (responsible for http, php, cgi etc.);
• MySQL server (responsible for work with MySQL databases);
• Mail server (responsible for e-mail services: POP, IMAP, SMTP);
• FTP server (responsible for FTP connections and transfers);
• SSH server (responsible for SSH jail/full shell connectivity);
• DNS server (handles domain name related operations and settings).

To use each of these services, you need to be connected to each one of them. If you are browsing through your website, you are likely using only Apache server (i.e. http). At this point you can already think of this process as the simplest http uptime check (by sending connection requests and receiving results from a server). If you are checking emails with a mail client, such as Mozilla Thunderbird or Microsoft Outlook, you are interacting directly with the mail server. When uploading new content using FTP client, you are only communicating with the FTP server. As you can see, in a variety of cases operations performed by a user at one given point in time include either one or few services simultaneously. Imagine a situation when you can successfully receive mail messages, but at the same time nothing (normally “nothing” is considered to be a certain kind of error, 401 for example) is displayed when you are viewing your web site via browser (i.e. a downtime situation takes place). This often means that Apache server is experiencing problems (as html pages are normally processed by the http service).

When we are considering availability, we should keep in mind that server downtime can be either full or partial. It is normal for a hosting company to guarantee a 99,9% uptime figures, as the underlying value  normally assumes percentage of time a web server is not spending in full downtime state. As any other devices, servers can and do get broken for various reasons. In many cases it is very hard to predict when exactly and why a server can fail. What is more important is whether a hosting company is capable of fixing the issue in time and with minimal impact on the customers. We at Sitevalley.com spend a lot of resources on server monitoring and always have a backup plan in case something goes wrong. We use special network monitoring software to constantly track status of our servers. Our network’s multiple redundant backbone connectivity links all of our Datacenters, using multiple high capacity lines. These Datacenters are maximum security, carrier-class facilities with fault tolerant power systems. All of them have state of the art IP architecture aimed at reliable delivery of your content 24x7x365.

Web site monitoring tools

After having defined uptime meaning and importance, let’s take a look at monitoring tools available for hosting users. We immediately can divide these checking utilities into two primary groups:

• free;
• paid.

The paid solutions mostly include the web based services that perform monitoring of web sites from numerous locations across the Globe. The technology behind such utilities is as follows: a certain number of interconnected servers with installed monitoring software send various requests (http, MySQL and DNS) to monitored locations. Based on the responses they receive, using special algorithms, they make conclusions about the current state of the location (i.e. a web site). Consequently, these results are stored and pictured as graphs with time dependence. Users utilizing such tools can be notified about critical states of their web sites via e-mails, instant messengers, SMS messages and so forth. The testing results posted at labnol.org will be very useful for those looking for a good paid server monitoring solutions. Main advantages of the paid uptime checking solutions are:

• consistency of results;
• higher accuracy due to numerous locations the testing is conducted from;
• high response speed;
• variety of notification methods;
• service quality;
• independence from your own computer and ISP;

The main drawback is price, of course.

Free website monitoring services are normally divided into the following three classes:

• standalone server applications (need to be installed on a web server);
• free monitoring services analogous to the paid ones;
• standalone desktop software.

Free monitoring web based services use the same technology as paid services, but they offer lower response times and lower accuracy of results. The latter is due to the fact that such services normally use less monitoring points (i.e. locations) across the Globe to perform checks. There are exceptions, of course, one of which being a GIGRIB service, offered by the Pingdom Company (a very popular and effective monitoring services provider). GIGRIB uses its own customers as monitoring sources (this statement refers to the PCs of the customers). Such approach provides a very high accuracy due to high number of users.

Standalone server applications require installation on a separate hardware server. This type of solutions is widely used by webhosting companies to perform in house network monitoring. At the same time, standalone desktop applications are the worst and least efficient choices for users. These uptime monitoring techniques are very limited by the local ISP’s service quality. If the ISP servers go down, then your webhost will appear as unavailable for the applications relying on the local network. Such alerts may lead to false evaluation of your host’s service quality. You can find several lists of popular services from all presented categories in the following articles:

• Monitor Website Up / Downtime: 30+ Free Web Services and Tools
• 13 Website Monitoring Tools To Help Prevent Downtime
• Top 10 free website uptime monitoring services

Conclusions

Web site uptime monitoring is crucial for those users, whose income directly depends on the availability of their content. Uptime figures themselves are a very complex notion, dependent on various parameters, and when evaluating them, one should always keep in mind diversity of the conditions that influence these values. It is important to understand, that there are various types of uptime guarantees used by hosting companies. These guarantees very often assume the partial uptime figures. Monitoring availability of your own hosting account or server is quite important. Timely detection and reaction to the downtimes may help you save money and reputation. The descriptions and links given in the article will help those seeking proper uptime control utilities find them. This should consequently lead to higher income and ratings of web resources.

]]> http://svhostingblog.com/technology/web-host-uptime-monitoring-purpose-and-tools/feed/ 9 http://svhostingblog.com/technology/xen-and-openvz-technology-insight-and-comparison/ http://svhostingblog.com/technology/xen-and-openvz-technology-insight-and-comparison/#comments Tue, 08 Dec 2009 16:16:16 +0000 Smirnovi4 http://svhostingblog.com/?p=329 Tweet

Virtualization is a widely used technology nowadays. A whole set of hosting plans is based on it – the so-called Virtual Private Servers (VPS). They allow steady transition from regular shared to the most powerful dedicated solutions. While big projects may require the power of an independent dedicated server, some personal and small-to-medium businesses may not need such resources at high costs at first. At the same time the needs of  such customers may not be satisfied with what regular shared hosting has to offer due its nature . Unlike shared hosting, V PS allows full isolation from other users on the host server. It provides full control over the account (i.e. root access), remote reboots and system restore.

Nowadays various approaches and implementations of virtualization exist. In this article we will compare two most widely used virtualization engines: OpenVZ and Xen. The main goal of the article is to provide basic concept, and outline the differences and similarities of the two engines.

Virtualization

As defined by Wikipedia virtualization is a term that refers to the abstraction of computer resources. In case of VPS  hosting plans, platform virtualization is used. Its idea is to separate an operating system (OS) from hardware it is being run on. With no virtualization applied, normally only one operating system can be run on one set of hardware at the same time. As depicted in the Figure 1, every server composed of a definite hardware set can simultaneously run only one OS, however, if the virtualization technology is applied, one achieves the ability to have numerous OSs run on single set of hardware at the same time:

Figure 1. Application of virtualization. Source: http://mysite.stonybrook.edu/

As we can see, the virtualization is used to create an independence from the hardware for an OS. Virtualization software not only allows to run several OSs on the hardware, but also separates the resources a single OS is consuming. This way various quotas for CPU, RAM, HDD I/O and such can be defined for a single OS. Quota based isolation increases the utilization of a single dedicated server by dividing resources of the latter into smaller chunks. This is quite efficient as it decreases the cost of a single VPS and, at the same time, provides new unique features to the user.

Paravirtualization

In order to understand how virtualization works let’s consider its structure (see  Figure 2 ). As we can see form the figure, main idea of virtualization is creating a special interaction layer (Virtualization layer) between the hardware and OS which acts as a resource isolator and manager at the same time:

Figure 2 Basic structure of virtualization. Source: http://desktop-virtualization.com

Figure 2 shows a special case when the virtualization layer is placed directly on top of the hardware (architecture) and then the OSs are installed on top of the virtualization layer. Consequently, the applications such as Apache Web Server and MySQL server are run on top of the OS and do not interact with the applications run on the other OSs. This way the depletion of resources given to one OS does not influence the performance of the other OSs on the hardware carrier server. Such isolation eliminates the possibility of overloads caused by other users on the carrier. This kind of virtualization can be referred to as to a Paravitualization.

As defined by Wikipedia paravirtualization is a virtualization technique that presents a software interface to virtual machines that is similar but not identical to that of the underlying hardware. This technique is used in the XEN virtual engine. Paravirtualization has a number of limitations, one of which being a special need for slight modification of the OS kernel in order to teach it to use the instructions given by virtualization layer. Such need, however, can be avoided by utilization of special processors supporting the hardware assisted virtualization technology. A processor with such capability can help the software virtualization layer to perform the abstraction. This way no modification is required to be performed to the OS, and therefore any proprietary software like Microsoft Windows can be run on the virtualization layer. Not only it allows for running of proprietary technologies, but it also lets using the original form of the OS which can be important for a number of applications run on it.

Operating system-level virtualization

Paravirtualization is not the only type of the virtualization used in hosting industry. Another widely used type is Operating system-level virtualization (OSLV). This kind is defined In Wikipedia as a method where the operating system allows for multiple user-space instances (virtual hosting, chroot jail + resource management). This means ,that only one OS with only one kernel is running on the server. OS is modified in such a way that it allows for creation of so-called containers which appear as a separate OS entity to the user and software within that container. This can be depicted as shown in the Figure 3:

Figure 3 Basic structure of Operating system-level virtualization. Source: http://www.linux-watch.com

This type of virtualization is implemented in the OpenVZ engine. Resource sharing in such structure differs from the paravirtualization. In the latter technology the resources are firmly distributed on the virtualization level, and no sharing takes place. In the OSLV all containers (VPSs) are running in the same environment. The resources of each one are assigned and limited. However, there is a certain amount of the resources that is kept for sharing within the host OS and is not assigned to any specific VPS. If such resources are not occupied by host system, any VPS can use them if necessary. An example can be given in terms of a burstable RAM. There is a guaranteed RAM quota VPS always has, and there is also a top limit of burstable RAM this VPS can also use in case the system has any additional free RAM besides the amount already guaranteed and provided.

In OSLV all containers run inside a single host system. Each container emulates an OS within itself, but it can only emulate the OS kernel which is similar to the one the container itself is running in. This is a great limitation. OpenVZ is run on Linux OSs. Therefore, the only OS that can be installed in the containers within the carrier is of Linux. This way usage of FreeBSD is simply not possible within the OpenVZ based VPS. At the same time, there is no problem using FreeBSD in a paravirtualized environment like Xen.

Xen vs OpenVZ

Xen

Figure 4 Xen paravirtualization architecture. Source: http://uberhip.com

Xen is a virtualization engine (to be exact it is a virtual-machine monitor) for x86, x86-64, Itanium and PowerPC platforms. On number of processor a paravirtualization technique is applied by Xen. This means that the operation systems run on Xen are modified in order to achieve high performance on a wide range of hardware architectures, which are initially not intended for virtualization technologies. The architecture of Xen is presented in Figure 4.

Xen uses one privileged OS to control the hardware resources. This privileged OS must be a Unix based one (e.g. Linux, FreeBSD etc.). In the latest versions of Xen the following OSs can be run as guest ones:Linux, Minix, Plan 9 from Bell Labs, NetBSD, OpenBSD, FreeBSD, OpenSolaris, NetWare, GNU/Hurd/Mach, OZONE.

OpenVZ

Figure 5 OpenVZ operating system-level virtualization architecture. Source: http://www.linux-watch.com

OpenVZ is an open-source virtualization engine on the x86, x86_64, and IA64 processors. OpenVZ, itself, is built on top of Linux. Unlike Xen’s paravirtualization technique, with OSLV virtualization the operating environment is virtualized instead of the hardware. Thus, while there is only one operating system kernel, multiple programs run in isolation from each other within the single OS instance. The basic architecture of OpenVZ is shown in the Figure 5. OpenVZ project manager Kir Kolyshkin said in a statement about its architecture:

“Unlike hypervisor-based virtualization technology [such as Xen and VMware], OpenVZ is extremely easily portable across different architectures, since 95 percent of the code is platform-independent. With the port to Power, we’ve opened the door for another segment of the user community to benefit from our open source virtualization software”.

Differences and Similarities.

Both system present the user with the following benefits:

• consolidation
• increased utilization
• rapid provisioning
• dynamic fault tolerance against software failures (through rapid bootstrapping or rebooting)
• hardware fault tolerance (through migration of a virtual machine to different hardware)

Resources

Additionally both Xen and OpenVZ provide the the ability to securely separate virtual operating systems. However, this feature is implemented through different techniques in these two virtualization solutions. Xen provides full fixed isolation, where the initially assigned quotas are kept throughout the entire functioning period, and additional request for memory are processed using the swap space on the HDDs. OpenVZ ,on the other hand, provides semi-dynamical assignment of resources. Such terms as burstable RAM and resources exist on OpenVZ. Therefore, unlike the case with Xen, VPSs based on OpenVZ may also take advantage of the free resources of the server. This can result in better utilization of the resources of the carrier hardware.

Performance

The virtualization overhead observed in both OpenVZ and Xen is limited. Various opinions exist on the difference in performance between the two. However, in both cases the performance levels of virtualized environment, as compared to the real hardware, are of acceptance-quality level. Specific figures depend on a great number of factors and cannot be summed up for the general conclusion.

Compatibility

Unlike OpenVZ, XEN has the ability to support legacy software as well as new OS instances on the same computer. That means that proprietary systems can be installed on Xen based carrier without any additional modification if hardware assisted virtualization is used. OpenVZ provides compatibility only in frame of the alike kernel, such as various distributions of Linux OSs.

Migration

Both provide great and relatively equal check pointing and live migration options.

Scalability

Both engines are based on the Unix OSs, therefore they have great scalability. For example in case of OpenVZ which employs a single kernel model, it is as scalable as the Linux kernel. Such kernel supports up to 64 CPUs and up to 64 GB of RAM. (on 32-bit with PAE) A single container can scale up to the whole physical system, i.e. use all the CPUs and all the RAM.

Management

Peculiarity of OpenVZ is the mass management support. An administrator (i.e. root) of an OpenVZ physical server (carrier) can see all the running processes and files of all the containers on the system. That makes mass management scenarios possible. Let’s look into the case in which  Xen is used for server consolidation: in order to apply a security update to 10 virtual servers, an administrator is required to log in into each one and run an update procedure. Since VPSs are normally indented for different users with their own projects, this is not an issue for Xen based virtualization. However, the fact that with OpenVZ a simple shell script can update all containers at once greatly eases the management of the entire system.

Conclusions

Virtualization basics as well as the main principles of two fundamentally different virtualization solutions have been described in this post. It is obvious that both projects are outstanding virtualization systems and provide host industry with sufficient number of features to support an entire type of VPS plans. Each of the engines is aimed at solving specific set of tasks, which they successfully do. Xen is a technology mainly aimed at providing the maximum approach to the full virtualization and full separation of virtual machines on the hardware carrier. It is also aimed at supporting the maximum range of existing operating systems. OpenVZ, on the other hand, is a system specifically aimed at Linux based virtual environments. It provides good dynamic resource sharing and isolation tools, which leads to higher server utilization figures. Both engines show great performance.