We received an urgent call from a business partner. One of their customers had fire in a Rack in their Data Center and they were looking for a solution to prevent a recurrence.
When our engineer visited them, he had two surprises. One, they had absolutely no monitoring system in their Data Center. Two, they had no clue what caused the fire. He inspected the Rack and it appeared overloaded considering its power density. No one in the team could provide details of the devices placed on that Rack. They did not even have a spreadsheet!
There were no rack or aisle temperature sensors. Just couple of room thermometers that were manually monitored. The good news was they were working!
Of course, they needed to do something! At least start with a spreadsheet. But for the large majority that would be woefully inadequate. An automated Power & environment monitoring system providing alerts would be a bare minimum. DCIM would help!
For a large part, DCIM has been considered a Facilities play, as optimization of key Data Center resources – power, space and cooling. However, a modular DCIM supporting IT Infrastructure Management could prevent fire or power trip in a Rack. In an environment operating without accepted Data Center Standard Operating Procedures (SOPs), instruction to add few IT devices in Racks would typically be done with ad hoc placement on vacant U-spaces. That usually is the cause of unfortunate incidents.
DCIM – built with SOPs – would have Workflow-based Move-Add-Change operations that would first identify best-fit Rack based on its capacity versus utilization before an approval is granted. Provided a Rack has iPDU or there’s branch circuit monitoring, we would get actual power consumption on the rack. If nearing, allocated power, we would avoid adding a HPC server to avoid a potential power trip. Similarly, if we have temperature sensors on the rack and it’s showing near upper tolerance level, we would again avoid that rack. We could also estimate weight of the Rack if the DCIM has a comprehensive OEM library that includes U-space and weight of the IT devices, and this information is populated for each Rack.
Most DCIM add interactive graphics to get all this information (power, space, heat load, weight) in one-go to identify best-fit racks for adding new devices on racks. It can avoid costly mistakes.