SAN Management Buyer's Guide
Despite the immense changes in IT over the last decade, the Storage Area Network (SAN) remains the backbone of the storage world. Indeed, the complexity of modern environments has grown to the point where SAN management tools have become a necessity, particularly in larger enterprises.
Here is a sampling of what is offered in the SAN management space.
Brocade has been the big name in SAN management since the early days of Fibre Channel (FC). The company claims to have an installed customer base encompassing 90 percent of Fortune 1000 companies, using Brocade Fabric Vision. This is a hardware and software combo has been updated with Monitoring and Alerting Policy Suite (MAPS) and Flow Vision features aimed at simplifying SAN administration, boosting SAN performance and reducing costs.
MAPS monitors SAN health and alerts operators on situations inhibiting uptime and availability. Threshold-based rules and policies can run across an entire SAN fabric or multiple fabrics. Users can customize rules for specific ports. Flow Vision, on the other hand, is all about visibility into flows. Users can obtain statistics associated with specific flows concerning application performance. Together with MAPS and Brocade Network Advisor, Flow Vision is said to enable more proactive application management.
“MAPS was developed because customers told us that activities such as monitoring and setting thresholds for their SANs was complex, cumbersome and time consuming,” said Scott Shimomura, Director, Product Marketing, Brocade. “It distills more than 170 manual steps and more than 15 years of best practices into one mouse click. It helps eliminate more than 50 percent of common issues such as CRC errors, invalid transmission words, link resets and loss of synchronization.”
SolarWinds Storage Manager provides heterogeneous storage management, including single pane management for arrays from EMC, NetApp, HP, IBM, Dell and Hitachi Data Systems (HDS) among others. It comes with agentless DAS, NAS and SAN performance monitoring, as well as generating policy-based alerts on status and usage thresholds. In addition, it offers capacity planning, physical to virtual storage mapping, file analysis and storage I/O hotspot detection. It is priced by the number of disks managed, starting at $2,995 for a perpetual license for up to 25 disks (includes the first year of support). There is also volume pricing. For example, list price for 500 disks is $23,990.
This product is mostly going after the multi-array market. Those not so likely to use Storage Manager are companies with environments with only a single storage array vendor. In that case, the tools provided by the vendor are adequate for their needs, said Michael Thompson, Principal, Product Market Management, SolarWinds.
“Storage Manager gives admins the ability to look at multiple layers and numerous performance counters to allow for identification of performance trending and bottle necks,” said Thompson. It also helps the find unused or allocated but free storage on arrays.”
Virtual Wisdom by Virtual Instruments monitors, reports, trends and diagnoses infrastructure performance issues in a vendor agnostic way. It does this through a combination of software and hardware. It is comprised of several elements such as SAN Availability Probe that gathers status from SAN switches via SNMP. It achieves this by collecting metrics and fault conditions for each port (such as the number of frames and bytes, loss of synchronization, link resets, link failures, packet discards, and CRC errors). Further software probes monitor servers and all aspects of SAN performance. They link up with Virtual Wisdom servers and dashboards, as well as its SANInsight hardware to provide visibility into heterogeneous virtual infrastructures.
Manage Engine’s Op Stor is another multi-vendor tool. It monitors SAN and network attached storage (NAS) devices such as storage arrays, fabric switches, tape libraries, host servers and host bus adapters cards from a multitude of vendors. It performs auto-discovery to deliver asset, capacity, performance, and configuration information on storage devices and their components. Users can model their SANs to optimize the configuration or debug performance.
Armed with up-to-date device data, administrators can key in details such as device location, vendor details, and technical support info to stay in control of their SAN. This feature is also of value when it comes to device-level problem reporting and maintenance. Additionally, Op Stor captures sensor faults such as power supply, temperature, fan and battery, as well as status change notifications. Administrators receive alarms through email or SMS messages based on pre-set rules. Finally, it helps identify performance bottlenecks by analyzing historic and current usage trends for storage capacity along with traffic utilization trends to generate forecast reports.
Cisco Prime Data Center Network Manager (DCNM) provides centralized management of storage arrays, servers and switches on Ethernet and FC networks. It gives a converged view across platforms running the Cisco NX-OS operating system, which is used by the Cisco Nexus Data Center Switch and Cisco MDS 9000 Multilayer Switch families. For SAN management, DCNM is said to simplify configuration and deployment. It has automated health check reports, health indexes, configuration analysis and host path redundancy checking.
Richard Darnielle, director of marketing, Cisco Data Center Solutions, said the user base ranges from those with small SANs all the way up to huge networks comprised federated DCNM servers where hundreds of switches with hundreds of fabrics support tens of thousands of SAN (and FCoE) ports. In particular, Cisco customers utilizing “director class” Cisco SAN switches are big users of DCNM. It is offered in a free basic edition as well as licensed for full SAN management capabilities, with list price of $7,000 to $21,000 per managed device.
“End-to-end visibility and management enables SAN operators network administrators and application owners to understand the relationship, path, health and performance from an application server (host) all the way through the network to and into the storage array,” said Darnielle.
Photo courtesy of Shutterstock.