Guaranteeing mission-critical applications using end-to-end infrastructure performance management

Customer Story and Case Study

Company Background

ByteLife is a data infrastructure company that builds, manages and guarantees critical business information systems. Its clients are companies whose latency sensitive transactions must be guaranteed against performance degradation or loss, which means that their data centers need to be 100% available. ByteLife’s main competencies are built around data centers: automation, management, virtualization, storage, networking, backup and archiving. Its primary technology partners are EMC, Cisco, VMware, VCE and Virtual Instruments. ByteLife customers are leading banks and telecommunication companies in the Nordic region and Baltic states such as Swedbank, Danske Bank, SEB, First Data, ABLV Bank, and the TeliaSonera AB companies EMT, Elion, Omnitel, TEO, and LMT. Founded in 2005, ByteLife has offices in the three Baltic states and Sweden. The company’s 2012 revenue was more than 10 million euros.

Customer Challenge

ByteLife recognizes that as the demands and complexities placed on data centers by mission-critical applications continue to increase, so does the need to guarantee their performance and availability. To show customers how to optimize performance and availability in a real-world setting, Tarmo Katmuk, ByteLife solutions architect, set out to create a small-scale version of an actual data center with two separate, but connected, data centers running applications independently.

ByteLife's IT Environment

The storage environment in the two ByteLife data centers includes EMC® VNX5300™ unified storage arrays with approximately 20 terabytes capacity. EMC® Fully Automated Storage Tiering for Virtual Pools (FAST VP) enables ByteLife to automate tiering among various drive types from SATA to Fibre Channel to Flash Drives. Fast Cache is used to speed up I/O response times within the storage array. EMC VPLEX virtual storage technology is installed in both data centers and provides application mobility and collaboration, ensuring availability during planned and unplanned downtime scenarios. The host environment is FCoE connected to Cisco Nexus 5000, and VPLEX is a full Fibre Channel front end and back end solution. VPLEX is connected to MDS switches in both data centers with the VPLEX connection going through Cisco 10 gigabit Ethernet switches (there is a 10 gigabit link between VPLEXs). The data centers feature Cisco® UCS B-Series blade servers, and the hypervisor used is VMware ESX 5.1, which is more than enough for ByteLife’s test environment. The Oracle RAC environment—a complicated real application cluster environment— has been tested extensively.

The most significant challenge in deploying mission-critical, latency sensitive applications is guaranteeing the performance and availability of the underlying infrastructure when new components like storage virtualization are introduced. The need for end-to-end performance visibility requires a holistic approach that continuously measures and monitors variations in latency at a sub-second level. It was tricky for Tarmo to measure latency because each component in the data center chain had its own native monitoring or configuration tool, but there was no end-to-end view of the overall infrastructure. He attempted to resolve this by measuring latency, response time and I/O with and without VPLEX in the configuration and then comparing the delta.

The Virtual Instruments Solution

ByteLife established a partnership with Nicolas Kollerstedt, Virtual Instruments Regional Sales Manager, in 2012 when VirtualWisdom was added to its data centers to make them 100% available. The VirtualWisdom platform drives improved performance and availability against infrastructures supporting mission-critical applications, provides unbiased, real-time performance, health and utilization metrics, and delivers end-to-end visibility.

While ByteLife has a long-term partnership with EMC, Tarmo concluded that none of the EMC tools could monitor the holistic environment—from the storage layer up to the application layer and everything in between. He concluded that Virtual Instruments offers the only real-time monitoring solution that is simple to use while giving him an end-to-end view of the environment. Tarmo finds the time stamp functionality especially useful as it allows him to slide back and forth to get a historical picture of exactly what was happening in the infrastructure at a specific time.

Moreover, VirtualWisdom does not introduce additional complexity or latency into the environment. “If you go for an EMC VPLEX solution,” notes Tarmo, “I strongly suggest you have Virtual Instruments Traffic Access Points (TAPs) installed in front of VPLEX so customers can proactively monitor the environment and identify in real-time any issues that could cause latency.” For example, if the VPLEX system is overloaded, Tarmo can decide whether to add more capacity to the data center or move applications from one data center to another. He emphasizes that Virtual Instruments gives customers peace of mind through preventative maintenance and prescriptive guidance when they virtualize their storage using EMC VPLEX.

More importantly, VirtualWisdom’s end-to-end performance metrics enabled ByteLife to accelerate their VPLEX deployment, eliminate performance blind spots, and more accurately predict infrastructure needs. Tarmo and Nicolas plan to include TAPs in every major project they’re collaborating on that includes storage—VPLEX, VMAX, or whatever brand of storage they are working with.


Learn how VirtualWisdom can improve your IT infrastructure by visiting our website, or:

Request a Demo »