Akorri Inc. added a predictive analysis capability to its BalancePoint storage resource management software to...
detect issues that can lead to application outages, and also expanded its support for storage devices and hypervisors.
BalancePoint 3.5 includes a Predictor for Storage feature that identifies physical or workload I/O problems and their impact on applications. Akorri also added support for the Hewlett-Packard SVSP storage virtualization device, Hitachi Data Systems Dynamic Provisioning, Microsoft Hyper-V and NetApp V-Series.
"We've focused on disk groups to understand loads against disk to determine best performance," he said. "We provide early warnings from a storage perspective to prevent downtime and brownouts."
Predictor for Storage detects disk contention issues by identifying servers or applications exceeding their disk utilization thresholds. It sends administrators an email with detailed utilization metrics, showing which servers are negatively affected and the root cause for the problem.
Comstock said the Predictor module will be expanded next year to look at virtual machines (VMs), caching, CPU contention and port contention issues.
Bryan Snyder, senior system engineer at Pittsburgh-based Millennium Pharmacy Systems Inc., said the Predictor feature is the next logical step for BalancePoint.
"That's definitely one of the features I'm looking forward to," he said. "Akorri has all the data so it makes sense for it to tell me there's a SAN issue. It's hard to proactively predict when there's going to be an issue."
Snyder said he started using BalancePoint approximately six months ago to help manage a storage environment that includes an EMC Clariion CX3 array as well as VMware virtual machines. Before using BalancePoint, Snyder said he was frustrated because finding problems required a bevy of point products. He said BalancePoint reduces the time it takes to find a problem from two or three days to approximately 30 minutes.
"Before I started using BalancePoint, more likely than not I'd know we had an issue by somebody telling me 'My virtual machine is running slow,' or something else was running slow," Snyder said. "I had a number of tools I had to look at to figure out the root cause of a problem. I had an EMC storage tool. I had to look at [VMware] virtual center to see if there's anything in there. I might have to look at individual ESX servers, and I had to look at virtual machines with PerfMon or similar tools. I had to start at the virtual machine and work my way back up to storage, or start at storage and work to the virtual machine.
"BalancePoint does all of that data collection all the time," Snyder continued. "If somebody were to come in now and say 'I'm having a problem,' I can go into BalancePoint and do some sort of analysis much quicker."
Chris Wolf, a senior analyst at Burton Group, said Predictor for Storage is a step toward "true quality of service" for applications.
"Eventually, organizations want to assign true quality of service for given applications or virtual machines," Wolf said. "But as intermediate steps, they want to understand application workloads to plan capacity and reporting or tracking of anomalies. If I see an unplanned spike in traffic, I can alert the right folks to take action."
Akorri's Comstock said BalancePoint's real-time cross-domain analytics engine is a good fit for customers as they begin to move to a cloud environment because Akorri can look at an organization's entire infrastructure across physical and virtual resources.
"We can see our customers' end-to-end infrastructure," he said.
Burton Group's Wolf said for Akorri to achieve "cloud service nirvana," BalancePoint would have to pre-define quality of service levels for virtual machines. But that capability isn't even available from VMware now. "Akorri's at VMware's mercy as far as what the hypervisor can support," he said. "True quality of service at the VM level is not even at the hypervisor level now."
BalancePoint 3.5 is priced according to capacity under management, beginning at $8,750 for 5 TB.