We are pleased to announce the addition of Matt Steiner to the Backup and Recovery team here at Contegix! In his previous roles, Matt has been responsible for providing comprehensive computer services, onsite and remote technical support and system administration of Linux/UNIX and Windows servers. Matt graduated from the University of Missouri, where he studied Information Technology. Welcome aboard, Matt!

In his latest piece, ‘Welcome to the Hyper Wars,’ Sreedhar Kajeepeta explains that this disruption is driving a need for multihypervisor environment and management tools that will work with several hypervisors to allow businesses to harness the latest innovation of multiplicity. Additionally Kajeepeta explains, “If you can make a strong case for multiple hypervisors, budget for a hypervisor neutral management console that has support for unified security, backup, disaster recovery, and load balancing and related migration of VMs.”

At Contegix, this disruption was not MiraCloud^TM. Our desire to design a product that would enable our customers to decide for themselves which solution best suits their unique work loads – all inside one unified infrastructure for added convenience and freedom of choice – was prompted by their expressed desire for more options. Our platform delivers options and makes cloud computing easier for everyone, especially those who need a cohesive solution, with varying levels of storage, redundancy and management.

The industry is slowly recognizing the basic truth behind workloads for cloud computing, which is – workloads are not equal. Businesses are extremely diverse, even within the same industry. We are focused on empowering our customers through our innovation. We invite you all to look for additional features to MiraCloud rolling out in 2012 geared toward making our multihypervisor approach to cloud computing even easier for our customers, and, ultimately, their users.

The holidays are a hectic time for everyone, in our personal lives, as well as in our business lives. At Contegix, we are especially tuned into the needs of our e-Commerce clients, who, at this time of year, are hopefully experiencing increased spikes in website traffic due to online shoppers. Andrew, one of our Engineering Team Leads says, “There are several components to capacity management for e-Commerce sites, but the most important aspect is the partnership between ourselves and the customer. Without great communication and the ‘team’ approach, it would be much more difficult to manage.”

Indeed. The process of onboarding for our customers often includes evaluations conducted on the client side, but also on ours. We like to plan ahead for web traffic spikes as much as we can, which often results in temporarily adding to a client’s infrastructure to handle seasonal spikes. Andrew explains, “Many times we will spin up cloud instances to manage these spikes, adding additional nodes to the same load balancer to facilitate seamless communications between dedicated physical servers and virtual servers. It’s an exciting part of my job because creating these types of infrastructures means we are providing true hybrid hosting for our clients who need it.”

Knowing what your business needs are, is certainly helpful in the communication process with us, because it helps our technical engineers design appropriate infrastructures rather than selling you capacity that isn’t necessary. “We are most successful with a combination of measurement provided by our e-Commerce clients that include traffic data from previous years and predicted levels for the current year. On our end, we plan for ultimate flexibility, leveraging our cloud assets within MiraCloud^TM, we can spin up and scale back capacity based upon system performance,” said Andrew. “Our managed clients have the security of knowing that their sites are constantly monitored 24/7/365, which allows us to get out ahead of any bandwidth or throughput issues that could arise due to seasonal traffic spikes.”

Couchbase and Contegix would like to invite you to our upcoming Meetup in St. Louis on Nov. 7, 6:00-8:30pm. There is no charge to attend, AND pizza, drinks, and a t-shirt souvenir will be provided!

Matt Ingenthron of Couchbase will be there to discuss “Everything Couchbase: Server, Mobile and Modeling Apps.”

You can sign up for free here: http://couchbase-st-louis-meetup.eventbrite.com

The meetup is at HALE Communications located on 1099 Milwaukee St. in Kirkwood.

About Matt Ingenthron:

Matt heads up the Developer Solutions group for Couchbase and has years of background in developing the backend systems for many large-scale, consumer facing web applications. For Couchbase, Matt leads the team developing client libraries, samples and docs for developers.  In recent years, he has been a core developer on Membase and Couchbase, a contributor to the memcached project. He is a frequent speaker at technology conferences including JavaOne 2011, MySQL 2011, MySQL 2010, OSCON 2010, Erlang Factory, Cloud Camp San Francisco, Silicon Valley Code Camp, and many others.

About Couchbase:

Couchbase is one of the leading NoSQL database software projects, with code, binaries and community residing at www.couchbase.org.  Couchbase, powered by Apache CouchDB, Membase and memcached Open Source projects, runs on everything from large-scale clusters in cloud compute environments down to iOS and Android mobile devices.  Its primary development is backed by Couchbase, Inc.

Naysayers, be warned. Technology analysts say cloud computing is here for the long haul.

Forrester expects cloud computing will increase from approximately $41 billion this year to more than $240 billion in 2020.

Analysts from Gartner are seeing a similar trend, forecasting that the cloud services industry will make an estimated $177 billion in revenue by 2015.

As we look deep into the crystal ball, the makeup of a business workforce will continue to become more mobile. Globalization is accelerating the need for more secure data and more efficient IT systems. Applications are as likely to be delivered via Software as a Service (SaaS) as they are via desktops. And companies of all sizes need to be able to scale their businesses to meet real-time demands. All of these factors, among countless others, contribute to the exponential growth of cloud computing.

But there isn’t a one-size-fits-all solution when it comes to the cloud. Some companies want to maintain private clouds for their VPNs. Start-ups, SMBs and enterprises are all integrating SaaS applications at different rates, and few companies are willing to move ERP systems to the cloud. This means businesses need customization.

One of the benefits of cloud computing is that with it, companies can choose the most efficient architecture to meet their unique needs. Think of it like a mix and match, from colocation to completely managed by an external provider and anything in between. With cloud providers like Contegix, businesses can even choose which applications and areas of their business to move to the cloud and can select their level of service.

Regardless, of each company’s cloud migration strategy, there is no denying the growth of cloud computing. Now is a good time to set a path in the cloud. As the famous H.G Wells once said, “Adapt or perish.”

For the month of October, Contegix’s Data Center Technician, Vernon, agreed to discuss his quirks, his work and his passion for his job here at Contegix.

Vernon’s father gave him his first computer – the Commodore VIC-20 in 1986, when he was only 8 years old. Vernon explained, “My career in IT began with that gift. When you grow up with Information Technology, it is always there. This is true for everyone. Most kids today cannot imagine a time when cell phones were not owned by everyone; I have that attitude about computers.” It was with the Commodore that Vernon first experimented with writing code and manipulating technology to see what it was capable of accomplishing. At the time, technology was new to everyone. As time went on, Vernon embraced technology and valued the excitement felt when he could try out his new ideas using technology.

As our one of our Data Center Technicians, Vernon is responsible for physically racking, wiring, shelving, connecting and testing our equipment and is part of the team managing our facilities. He works meticulously to ensure that the presentation of equipment is not only functional, but also aesthetically pleasing. Again, his father, who was a sub-station mechanic for a power company, instilled this value in Vernon. He is absolutely meticulous about the architectural design of our facilities, to ensure that ultimately, our customers benefit from his clean and seamless layout designs. “My geek passion is centered upon starting with an empty room and boxes of parts only to end up with a beautifully laid out, well-organized data center,” explained Vernon. “I love creating order and functionality out of nothing. It’s what I was born to do.”

Vernon was one of the key architects in the physical layout design for MiraCloud™. MiraCloud incorporates a large amount of equipment in order to deliver our promise of a truly multi-tiered cloud platform. It is comprised of a variety of hosts, each with a different hypervisor (KVM, XenServer, or VMware ESXi). Each host needs access to the multiple storage units, to deliver primary and secondary storage. All of this presents the unique challenge of building something that serves the initial installation requirements and well beyond.
“The objective was to create a layout that would make future maintenance as easy as possible,” Vernon explained. “Functionality is about more than whether or not the servers fit into the allotted space and the proper amount of power is allocated to each unit. For us, functionality directly relates to maximizing uptime, elimination of potential human mistakes and creating a supreme level of accessibility, for our engineers and technicians who work on the infrastructure after the initial build out. We believe that this approach allows us to offer our clients additional confidence in the reliability of our systems.”

Vernon also takes his passion for technology and building things seriously in his personal life. Recently, he purchased a Honda Ruckus and modified the frame to suit his tall stature. In Vernon’s words, “I focus on function first, form second. There are plenty of examples of beautiful things that aren’t practical. My preference is for things to possess both qualities.”

When asked what Vernon enjoys most about working at Contegix, he found it tricky to pin down the right words and ended up saying things like, ‘culture’ and ‘balance.’ There is some difficulty associated with putting our company energy into words. We work in an industry that is all about innovation and forward thinking. Problem resolution in this business often means challenging conventional thinking. “At an early age, my dad taught me that just because things are done a certain way, it doesn’t mean that is the way those things should be done,” explained Vernon. In short, working at Contegix satisfies Vernon’s need to innovate, create practical beauty from chaos and contribute to the healthy balance of teamwork and individuality that makes Contegix such a great place to work.

Discover How MiraCloud^TM Can Help You Get Back to What You Do Best

We are excited and proud to announce that our new cloud platform offering – MiraCloud is now available. MiraCloud is designed for anyone who is searching for a powerful, flexible and customizable solution to address their unique cloud computing requirements. With a truly multi-tiered platform including options that range from self-service computing, to private enterprise cloud, coupled with the ability to choose elements from each tier for specific mission critical needs, MiraCloud is the first cloud offering of it’s kind. Making our customers’ lives easier is our top priority, which is what we challenge ourselves to achieve in all of our development projects. MiraCloud was born from the desire not only to reach this goal, but also to exceed well beyond it, and with all the features our offering provides, we are sure this platform will not disappoint.

We value innovation at Contegix, because in delivering advances on our technology offerings, we provide additional freedom to our clients. For example, MiraCloud offers three classes of service, including Convenience, Business and Enterprise Classes. To provide additional freedom, the three classes of MiraCloud can be combined in various configurations, allowing customers to mix and match the instances that work best for their specific needs.

With MiraCloud, our desire was to design a product that would enable the success and growth of our customers, who are constantly telling us that they need more options to meet the demands of their workloads. We know that the truth about workloads for cloud computing is that there is no equality, which creates an exciting challenge. Our tiered approach, including the ability to utilize all three tiers inside one unified infrastructure is meant to address this inequality. To learn more about the differences between the classes of MiraCloud, click here.

Within all tiers of MiraCloud, customers receive technical features such as operating system support; secure AJAX console access, a flexible pricing structure, dynamic scaling, custom ISO and a powerful API. We didn’t mess around with storage either; all MiraCloud instances provide persistent storage with additional storage on demand and volume snapshots. Network and Security features are covered in all tiers as well, with direct network and user VLANs, software load balancer integration, VPN support, software firewall integration and additional IPs on demand.

If you are interested in MiraCloud, but aren’t sure where to start, give us a call to discuss your needs with our experts at 877.4.CONTEGIX.

What does a sysadmin do exactly? The short answer is, nearly everything when it comes to the technology behind our personal and business lives. To demonstrate the scope of responsibilities attributed to sysadmins everywhere, here’s a short list: Server Installation, Network Configuration, Operating System Install, Server Temperature Control, Firewall Set Up and Maintenance, Email Configuration and User Set Up, Data Backup and Restoration – the list goes on. Sysadmins help users out when they need hardware and software support. Sysadmins monitor networks for problems. They plan and map out systems to discover more effective ways to manage our precious computing lives. They build infrastructures that are more resilient to today’s ever-changing, agile world; and in so doing; sysadmins reduce company cost, advance innovation and contribute to the overall growth of any given organization.

The vast majority of employees at Contegix are sysadmins. Here, they do not just manage, build and maintain our organization’s systems; they handle all of the above – and more – for our clients too. To say this is an important day for Contegix is an understatement the size of Godzilla. One day is not enough time to allow us to properly thank our team, so each year, we dedicate an entire week to celebrating our sysadmins. Be sure to check the blog next week, where we will post photos and detailed specifics of our planned surprises for our stellar team.

Contegix extends a special “thank you” to Mr. Ted Kekatos, for reminding those outside the IT industry exactly why the people in these positions are instrumental to our success, AND for being a system administrator.

Zach, one of our Technical Engineers, shares his thoughts on system tools

In the Linux world, we love having choices: choices about our kernel configurations; choices about our desktop environments and window managers; choices about which tools we want to utilise for various tasks. As Systems Administrators and Engineers, we heavily rely on the availability of logs regarding system performance (CPU utilisation, available memory, disk I/O, et cetera). Subsequently, we need to ensure that the tool we select for logging system information is tailored to the specific needs for each respective environment in which we work. Fortunately, we have a large pool of logging tools from which we can choose.

Though there are many such applications, we are going to focus on two prominent ones: sar and collectl. As some basic background information, sar (an acronym for ‘system activity reporter’) is arguably the “standard” system performance monitoring tool as it is provided with the widely-available SYSSTAT toolkit.  Sar not only monitors the aforementioned basic statistics, but is also capable of logging information regarding network activity for all interfaces, TTY device activity, process creation, and much more. Collectl is a similar monitor/logging utility that can be run interactively or as a daemon, allows for fine-grain tuning of data reports (even in sub-second intervals), and can format its output in a multitude of different ways.  So, which is the “correct” tool?  That’s a question that is largely dependent on your intended usage.  Below is some information regarding basic usage of each tool.

Sar

Sar relies on the system activity data collector (sadc), and essentially reports the information gathered by sadc.  Sar stores all of the data gathered over a 24-hour period in an individual file located at/var/log/sa/saXX, where XX is the two-digit day of the month (e.g., 02 or 19). As one might suspect (given that the full date is not listed) these logs rotate each month. Therefore, if saving the logs for more than a month-long period is pertinent, there should be a cron job that archives them to a different location.

Using sar is fairly straightforward if one simply wants to see all of the information contained in a particular log. To see the data that has been collected thus far in the current day, simply issue sar without any options. To see the logs from a particular day other than current, use the -f option. For example, a snippet of the provided data from the 19th of the current month might look like the following:

# sar -f /var/log/sa/sa19

06:20:01 PM       CPU     %user     %nice   %system   %iowait    %steal     %idle
06:30:02 PM       all      0.08      0.00      0.42      0.00      0.00     99.50
06:40:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
06:50:01 PM       all      0.08      0.00      0.43      0.00      0.00     99.48
07:00:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
07:10:01 PM       all      0.09      0.00      0.44      0.00      0.00     99.47
07:20:01 PM       all      0.09      0.00      0.43      0.01      0.00     99.48
07:30:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
07:40:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
07:50:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
08:00:01 PM       all      0.09      0.00      0.44      0.00      0.00     99.47
08:10:01 PM       all      0.09      0.00      0.44      0.00      0.00     99.47
08:20:01 PM       all      0.09      0.00      0.42      0.00      0.00     99.49
08:30:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
08:40:01 PM       all      0.08      0.00      0.43      0.00      0.00     99.49
08:50:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
09:00:01 PM       all      0.09      0.00      0.42      0.00      0.00     99.48
09:10:02 PM       all      0.09      0.00      0.44      0.00      0.00     99.47
09:20:01 PM       all      0.09      0.00      0.41      0.00      0.00     99.50
09:30:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
09:40:01 PM       all      0.09      0.00      0.44      0.00      0.00     99.47
09:50:01 PM       all      0.09      0.00      0.43      0.01      0.00     99.48
10:00:01 PM       all      0.09      0.00      0.42      0.00      0.00     99.48

10:00:01 PM       CPU     %user     %nice   %system   %iowait    %steal     %idle
10:10:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
10:20:01 PM       all      0.08      0.00      0.43      0.00      0.00     99.49
10:30:01 PM       all      0.09      0.00      0.44      0.00      0.00     99.48
10:40:01 PM       all      0.09      0.00      0.42      0.00      0.00     99.49
10:50:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
11:00:01 PM       all      0.09      0.00      0.43      0.01      0.00     99.48
11:10:01 PM       all      0.09      0.00      0.44      0.00      0.00     99.47
11:20:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
11:30:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
11:40:01 PM       all      0.08      0.00      0.43      0.00      0.00     99.48
11:50:01 PM       all      0.09      0.00      0.43      0.00      0.00     99.48
Average:          all      0.09      0.00      0.43      0.00      0.00     99.48

Without specifying any additional options, sar reports information regarding CPU utilisation, and stats related to I/O. If only data from a certain time span are needed, the -s and -e options can be used. Using the above example, one could get just the data from 08:40 until 09:20 by issuing:

# sar -f /var/log/sa/sa19 -s 08:40 -e 09:20
08:40:01 AM       CPU     %user     %nice   %system   %iowait    %steal     %idle
08:50:01 AM       all      0.09      0.00      0.43      0.00      0.00     99.48
09:00:01 AM       all      0.09      0.00      0.43      0.00      0.00     99.48
09:10:01 AM       all      0.09      0.00      0.43      0.00      0.00     99.48
Average:          all      0.09      0.00      0.43      0.00      0.00     99.48

Though all of the options can be found in the man page for sar (man sar), one more important option is -A, which shows an incredibly large amount of additional information including but not limited to I/O, paging stats, block device activity, queue length, load activity, used cache entries, and used file handles.

Collectl

Collectl offers similar functionality to sar, but also includes a few additional metrics and configuration options. In order to simply start collectl in interactive mode and have it start logging immediately, just issuecollectl at the prompt. After a few seconds, data similar to the following will appear:

# collectl

#<--------CPU--------><-----------Disks-----------><-----------Network---------->
#cpu sys inter  ctxsw KBRead  Reads  KBWrit Writes netKBi pkt-in  netKBo pkt-out
  37  37   382    188      0      0   27144    254     45     68       3      21
  25  25   366    180     20      4   31280    296      0      1       0       0
  25  25   368    183      0      0   31720    275      2     20       0       1

and the stats will continue to be piped to stdout.

Like sar, collectl has myriad options that can be passed to it in order to manipulate the types and format of reported data. Collectl can also be started in the same manner as any other service, using service collectl start or /etc/init.d/collectl start, and it will run in the background. One of the most interesting options is -p or -plot, which generates output in a format that is space delimited, which works nicely with plotting applications like gnuplot. In addition to all of the information sar collects, collectl is able to gather information about many other subsystems by appending them with the -s option. For instance, to see data regarding NFS and Lustre, pass the option (using +/- format) -s+fi (+ to add a subsystem, f for NFS, and i for Lustre).

Conclusion

Now that you have seen a very brief introduction to two of the more prominent system performance loggers, which one should be used on a regular basis? The short answer is both. The long answer obligates one to investigate each tool extensively, and utilise each one for various situations. Both are lightweight applications, and offer incredibly detailed data regarding system health. Sar comes standard with RHEL and derivatives, while collectl will either need to be installed from within the distribution’s package manager or compiled from source. Collectl offers multiple polling intervals not available from within sar (without custom wrappers), but can be a bit daunting for basic monitoring. In Linux, we love our freedom of choice; do your research, make an educated decision about which utility best fits your needs for a particular situation, and go for it!

Here’s a clip from COSE Update article:

With the cloud acting as your server or your software, you always have uptime. You get sophisticated applications and a smart infrastructure. You get the stuff  you maybe couldn’t afford when you were upgrading servers or other computer hardware and software…

“At the end of the day, the questions are, ‘What does this do for my business?’ and ‘How do I drive value from cloud computing compared to the way I was doing things yesterday or 10 minutes ago?’” said Matthew Porter, the chief executive.

The mental barrier that’s still tough to grasp:  What’s up there? What’s in the cloud? “You see a commercial with a barista and all of a sudden he has three friends and a big business because they all went to the cloud,” he said.

Matthew went on to explain that choosing a cloud provider should be the same process as choosing any other vendor. Making sure your vendor is committed to your cloud infrastructure and works around the clock to make sure your data is safe is key to a successful relationship. Understanding the response time of the provider is the best indicator of the level of service a business can expect.

Read the full article here