Skip Navigation

Engineering Blog

Archive for January, 2010

SQL Enterprise coming to Cloud Enterprise

Tuesday, January 26th, 2010

From: Matt Ferrari, Director of Platform Engineering

A few months ago Hosting.com launched the ability for virtual machines within our Cloud Enterprise environment to support up to 8 vCPUs and 32 gigabytes of RAM per instance.  While that in itself was exciting, we did not have many customers that needed 32 gigabytes of memory for a single application driven environment.   We did, however, have many customers that push large SQL environments to the limit so our software requirements for the product expanded.

In the very near future we’ll be announcing the availability of SQL Standard 2005/2008 and SQL Enterprise 2005/2008 within our Cloud Enterprise environment.  Currently we’re running through the final paces to ensure stability, provisioning automation, minimum requirement rules through our control panel, and accurate license reporting.  New and existing customers will be able to login to their control panel and automatically provision SQL Standard or Enterprise into their new virtual machine.   Our engineering groups have some other technology announcements coming within the next week so be sure to stay tuned.

Posted in Platform Engineering | Comments Off

Don’t be shy!

Tuesday, January 19th, 2010

From: Adam C. Greenfield, User Experience Manager

Today I had the opportunity to speak to one of our customers regarding our Customer Portal and it’s future. This particular customer was a customer who originally signed up with HostMySite. He has a few dedicated servers and reached out his sales person and indicated an interest in discussing our Portal.

In this situation the customer was in the process of merging multiple organizations (much like we have been recently) and ended up having the account originally opened with us and one account with one of our major competitors. They had discussed things and decided they wanted to stick with us (primarily based on our support according to the customer). However, one of the concerns that was raised during their internal dialog was a number of things they would really like to see in our Customer Portal.

After Doug was introduced to me by his sales representative we scheduled some time to talk on the phone. Doug’s concerns mainly revolved around centralizing all of the information we send to customers during the set up process and various upgrades in to one single location. Luckily this was one of the major goals of our new Customer Portal, to create one single location for all interaction with our products.

It also turns out that Doug was still using one of our legacy control panels, and hadn’t really been exposed to our new Portal yet. Each one of Doug’s specific concerns were already things that we had scheduled for our engineering roadmap in the next quarter or two. For me, this was great feedback and a little validation for the direction we are moving in. Getting the chance to speak with Doug today was extremely beneficial for both of us.

I would love to hear from any customer (or prospective customer) that would like to discuss their thoughts about our portal, our services, or the customer experience and what they would like to see in the future. Please contact me (my first name at hosting dot com) and we will make time to talk.

Tags: , , ,
Posted in Software Engineering | 2 Comments »

Cloud Enterprise Architecture

Monday, January 18th, 2010

From: Matt Ferrari, Director of Platform Engineering

One of our first architectural action items that we pursued when designing our Cloud Enterprise environment in 2009 was to find great technology that we could integrate with.  The engineering groups took user requirements from our product managers and tested many different hardware and software platforms in order to find the best match to both meet our customers’ short term needs and long term goals.

While we have built many of the multi-site deployment automation and control panel functionality internally, our Cloud Enterprise environment uses technologies from many of our valued partners.  From our network switch and firewall side, we utilize Juniper Networks EX and SSG series to provide customers with the ability for isolated environments within Cloud Enterprise.  The hardware nodes themselves are Dell R-Series servers that utilize fibre channel connectivity to our fibre switches, which in turn connect to our SAN architecture in each Cloud Enterprise location.

For the SAN architecture we partner with EMC, which provides us both Fibre channel, SATA, and NAS options for our Cloud Enterprise environment.  On the hyper visor side (or your back end software environment) we deploy our nodes with VMWare ESX, now on VSphere 4.0.  Within that environment we utilize technologies such as High Availability and DRS in order to  provide our customer base with a high level of uptime.  In case of a resource failure we keep extra hardware nodes within the cluster so that the environment can continue to maintain availability in case of disaster.

As we continue to add features to that offering in 2010 we also plan to expand on the technology that the Cloud Enterprise environment is powered by.

Posted in Platform Engineering | Comments Off

Server Replication

Thursday, January 14th, 2010

From: Matt Ferrari, Director of Platform Engineering

One of the first reasons that we encouraged the creation of our new engineering blog was the ability to speak about some of the exciting enhancements that we are bringing to the market.  We had some initial concerns, including announcing a product that may end up changing in functionality before launch.  That being said the key driver here is better communication, both with our internal and external customers.  So I’ve decided to provide sneak peeks into some of the functionality that my group is preparing to launch in the upcoming future.

As we continue to grow our storage offering, one of the key components from a managed services perspective that we heard loud and clear from our customers was the need for host replication.  We have multiple data centers, multiple cloud infrastructures and dedicated networks, and even multiple SAN architectures across those environments.   Now many of  our customers want to be able to have a low RPO (Recovery Point Objective) and RTO (Recovery Time) replication service that in case of disaster they can avoid an unplanned outage.

In the near future, we will be launching a service that will provide Windows host to host replication across multiple data center infrastructures for just these disaster recovery purposes.  It will include the ability to perform tasks such as:

- Replication for Active Directory, Sharepoint, Microsoft SQL, Exchange, Web server, and other environments

- Highly Available Failover: Monitoring critical events via an agent that will provide the ability for global server and database service fail over, which can be set to either automatic or manual

- Scheduling of replication: Changes are replicated in real time as they occur or as scheduled

- Portal access for our client base to have a direct interface to the replication options, services, and health checks

We could go on, but let’s leave some of the feature excitement for the launch of the product.  From an engineering perspective we are working to test compatibility, stress levels,  fail over, initial deployment procedures, support, escalation, and portal compatibility.

Posted in Storage Engineering | Comments Off

Portal SDK: What does that mean?

Sunday, January 10th, 2010

From: Adam C. Greenfield, Lord of the Acronyms

During the past several weeks we have been in the process of creating a Software Development Kit for our new Customer Portal. This kit is targeted at internal engineering teams as well as trusted partners to enable the rapid development of new functionality for our customers that can be easily included within the portal.

As the portal represents a key part of many of our upcoming product offerings we want to ensure that engineers outside the core portal group have the ability to develop the customer-facing interface in tandem with the products themselves. Rather than being developed after the fact we believe that the portal components of product offerings should be treated as an integral part of the offerings and often the best way to do that will be to allow folks that know those products the best from top to bottom to develop the portal modules themselves.

In order to ensure the highest quality for our customers, the core portal team will still be serving as the final quality assurance and ultimately the arbiters for what makes it in to our production portal. We believe that this new Software Development Kit will aide our new portal to quickly become a powerful single pane of glass into all of our product offerings. Customers can look forward to a rapidly expanding menu of services made available within one easy to use location with a common look and feel.

Another concept this will open up is the potential for partners to white label our proven portal framework to expose functionality directly to their customers without having to undertake the work of starting from scratch. Enabling this sort of functionality has been an architecture goal since we originally began this development effort.

We have begun circulating this kit internally to engineering teams and hope to have it out to the first set of partners within a few weeks. So far there has been an unprecedented amount of interest and we look forward to adding many more services to our portal in the weeks and months to come. If you are a Hosting.com partner interested in our portal SDK (or would like to become one) please don’t hesitate to reach out to me. You can contact me at my first name at Hosting.com.

Tags: ,
Posted in Software Engineering | Comments Off

The Engineering Teams

Saturday, January 9th, 2010

From: Adam C. Greenfield, User Experience Manager

Over the past year we have made great strides in formalizing the structure of our engineering organization. We currently have three major groups within engineering. I wanted to take a moment to talk about each group and their primary area of focus.

The first group we’ll talk about is Network and Security Engineering. This team focuses on refining and expanding our critical network services. They are responsible for implementing products related to firewall, load balancing and serving as a resource to the rest of the organization when products involve specialized networking or security configurations. Maintaining a number of critical internal tools to support our operations team and our monitoring systems round out much of the day-to-day workload this team carries.

The next major group is our Platform Engineering team. Platform Engineering is our newest group and they work on a wide variety of systems architecture and host based products and solutions. During 2009 this team built our new Cloud platforms and the product sets surrounding them. In addition to the customer facing platforms this team oversees many of our automated provisioning and workflow automation tools.

Last but not least is the group I am proud to be a member of. Our Software Engineering group produces a number of web applications and tools. This includes our primary web site, our customer portal and our online ordering tools. Internally they are also responsible for our sales tools, inventory, billing and ticketing systems. Within Software Engineering there are project teams and a production support organization that deals with improvement of the wide range of internal and external tools used at Hosting.com.

Many of our strategic goals span the entire engineering organization and involve members from each of these teams. Along with our product managers and leadership these groups all focus on bringing new and improved offerings to our customers and staff. Today we have engineering staff spanning from East to West coast and are always looking to add more talented members to our team. If you are interested in joining the team at Hosting.com please don’t hesitate to check out the careers section of our website and submit an application.

Tags: ,
Posted in Uncategorized | Comments Off

Putting the Enterprise in Cloud

Friday, January 8th, 2010

From: Matt Ferrari, Director of Platform Engineering

When Hosting.com launched our first Cloud product in 2009, we didn’t expect the Cloud Enterprise product to evolve as quickly as it has.  Unlike a significant number of past product launches, our customer base quickly gave us feedback that they wanted more.  As a managed host, we’re proud of our Cloud Enterprise environment.  We provide the deployment, the managed operating system, the licensed software packages, and managed services that allow our customers to get far more than just a plain virtual machine out of the gate.

Technologies such as VMotion, DRS, resource growth on demand, API integration with vcloud, and other items have helped make the platform robust and built for business growth without painful migrations.  As our Platform Engineering team continues to develop and deploy strategic products and tools for the organization, we also work to bring some of the newest features available to our Cloud Enterprise customers.  An example of some of the recent releases specifically for Cloud Enterprise include:

  • ColdFusion 9 Enterprise
  • Up to 32 Gigabytes of memory and 8 cores of processing per Virtual Machine
  • Additional managed operating system support such as Red Hat Enterprise, Cent OS, Windows 2008 and 2003
  • Additional storage options to utilize fibre channel disk speed on the SAN
  • Cloud Enterprise available at a second data center; Louisville, KY
  • VMDK Upload Functionality: the ability to upload an ESX created VMDK from another environment into our Cloud Enterprise environment

We’ve got some exciting items planned for the Cloud Enterprise environment in 2010.  I’ll be using this engineering blog to discuss this, and all of the other products, that we are actively working on and try to give some sneak peeks where I can.

Posted in Platform Engineering | Comments Off

Deploying our Customer Portal

Wednesday, January 6th, 2010

From: Adam C. Greenfield, User Experience Manager

One of the first tasks I worked on when I recently joined our Software Engineering group was managing the deployment of our new Customer Portal. All of our previous web properties have been deployed using dedicated servers within one of our datacenters.

From a security perspective, we deployed it behind a pair of highly available Juniper SSG-series firewalls. We also leveraged our F5 Big-IP Local Traffic Manager to split load between multiple web servers, allow for future scaling, and ensure a highly available application for our customers.

What we did differently this time around is rather than leveraging dedicated servers as we had in the past, we decided to deploy to multiple virtual machines in our Cloud. This decision was made for several reasons, primarily:

  • Ability to add capacity on demand by adding virtual machines
  • Flexibility to upgrade resources to existing infrastructure without downtime
  • Opportunity to leverage the same firewalls and load balancing available to us with dedicated servers

It also made each of our virtual machines highly available on their own, isolating us further from outage or performance degradation in the event of hardware failure.

We also decided to leverage our EMC Celerra NAS to store file cache and other data shared between the multiple web servers and services servers. This storage is presented to our web servers as a CIFS/Samba share which can be accessed and updated by multiple nodes simultaneously without the need for a more complex clustering solution.

Not only is our new Customer Portal the latest and greatest offering from Hosting.com in the way of customer experience, it is also powered by the same cutting edge infrastructure offered to each and every one of our customers.

Tags: , ,
Posted in Software Engineering | 4 Comments »