Harvest Your Data with DCIM Software

Shekhar Dasgupta<br/>Greenfield SoftwareShekhar Dasgupta
Greenfield Software

Shekhar Dasgupta is the Founder of Greenfield Software.

While data centers manage petabytes of business or customer data, the data from most data center operations is either trashed or simply ignored. Add to that the fact that some data centers have no monitoring systems in place and, therefore, little data exist from their data center operations, and therein lies the irony.

While DCIM software helps to improve availability, we also derive significant insights about the operating conditions of the data center through detailed analysis of data captured in the DCIM. This matters because DCIM analytics can yield significant savings in operating and capital costs of the data center.

Let us see what kind of data is generated out of data center operations and the lifecycle of this data. The first set comes from the IT systems: servers, storage and networks. The second set comes from the physical infrastructure: power systems, cooling systems, smoke detection and fire prevention systems.

IT systems generate data such as CPU and memory utilization of servers, Input-Output operations per second (IOPS) of storage, bandwidth, and latency of networks. Data centers with more advanced operations capture power consumption, temperature and airflow from these devices. Physical infrastructure generates data such as power consumption, and if the right systems are in place, temperature, humidity and air quality. While such data is generated, it needs to be monitored. DCIM is one of the tools used for data center monitoring.

The most common but rudimentary reason for data center monitoring is to provide alerts. It is like rainfall monitoring. Just as trending rainfall data over time along with other weather-related data allows predictions about the quality of monsoon in the coming years, data center monitoring allows us to predict an immediate failure and take actions to prevent a catastrophe. If we know that a UPS failure can cause fire, we can isolate that UPS or shut down the IT equipment connected to it if we get danger signals in advance through real-time monitoring of the UPS. DCIM’s earliest adoptions came from data center managers that required a single monitoring system for their physical and IT infrastructure.

Where the purpose of monitoring is to only get alerts, the value of data is transient and mostly purged within a month. This is rare among DCIM users, however, as they realize their DCIM is a gold mine.

DCIM software users have taken that leap forward to retain and analyze the massive amounts of data captured real-time. As repository of data captured from every single device in the data center – from both IT and physical infrastructure – DCIM lends itself to deep analytics that can help data center managers make major cost saving decisions. A couple of examples, include: Removing ghost servers and increasing rack density can improve space utilization which can defer the need to build or rent more data center capacity; and increasing temperature in certain zones of the data center, where servers have lower utilization in certain periods, can reduce cooling costs.

DCIM analytics is like rain water harvesting that goes beyond monitoring. Just as rain water harvesting is a vital way to conserve water, DCIM has become the de facto platform for reducing data center operations’ costs as well as improving availability to higher levels through correlations of data coming from multiple devices.

Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.

Add Your Comments

  • (will not be published)