When SANs emerged, one much-touted benefit was that all enterprise servers would use a central storage facility, with the capability of increasing or decreasing the amount of storage available to each server.
For most networks, the reality is quite different. Each server has an internal drive, many have an internal or external SCSI array, and some may also have external Fibre Channel attached storage. That mix of platforms and devices tangles up effective storage and SAN management.
Softek Storage Manager 2.5.1 does an excellent job of loosening the knot, offering a platform- and vendor-agnostic management tool that works with virtually all major server and storage devices.
It provides an integrated console with several modules that address management and optimization of storage across multiple servers. Additional Performance Tuner, Storage Infrastructure Manager (SANView 5.5), Data Migrator, and Data Replication Manager modules manage storage on servers running Windows, Linux (including IBM zSeries mainframes), NetWare, HP-UP, AIX, and Solaris, as well as most SAN devices.
Aiming for Optimal
I tested Storage Manager at Softek’s Sunnyvale lab because it has a wide variety of storage devices and servers. The management console installed easily on Windows 2000 and 2003 Servers. New in this version is support for the Apache Web server, as well as IIS for browser access to data (IIS is still required for dynamic reports).
Each system to be monitored must have an agent installed. Log-in access is handled through Windows security, creating an admin user group and a read-only user group. Agents can be installed remotely through SMS for Windows or by remote mount for Unix and Linux; after they’re set up, agents self-update as needed. Storage Manager itself acts as the console.
One of Storage Manager’s strongest pieces is Space Optimizer, which shows all the storage attached to each server. Optimizer allows you to identify which storage volumes are full and which users are using the most space; it also can search for files types (such as MPEG or MP3) that you don’t want on your network.
Drilling down from server to storage device to specific file to user details is an easy and straightforward process. You can query by just about any topic, such as the 100 or 1,000 largest files on the system or files that haven’t been used for a year. Space Optimizer integrates with major backup utilities so it can automatically back up and delete unused files through the server’s backup program.
You can also specify how often Space Optimizer collects file information, depending on your application. Because each process collects data on each file, user, device, server, and switch on the storage network, you probably won’t want to run it too often.
The impact on the storage network is not onerous, however. Measurements are tiered, so data is collected in a local database on each server and the central management console runs queries on each server. Agents can be turned off automatically when not in use.
Space Manager can also collect raw data from the OS on disk utilization, using OS-level reports. On Windows, it collects about 40 individual metrics through the Windows reporting tool.
Actions are not limited to identifying and deleting files. The console’s “action sets” area lets you define and test rules to find files. It has GUI tools to include or exclude directory trees, file types, time since the file was last modified or last accessed, and more. Action scripting can work with any command-line command to migrate data, back up data, or use utilities that aren’t integrated directly with Space Manager.
The Chargeback module identifies users by the amount of space used, the volume of data changed or downloaded in a day, or other parameters, allowing the billing of end-users or groups. Given that more datacenters are moving to a business-unit model — rather than simply being cost centers — this may prove quite useful.
The optional Performance Tuner module is a cost-effective investment for large SANs. It collects data on throughput, availability, and responsiveness of storage for all the services in the enterprise, all tracked on a per-server basis.
Performance Tuner’s data also aids in identifying trends and peak usage times, so you can project where more storage will be needed or troubleshoot choke points.
Historical data correlation lets you compare information from two servers to see if they’re both showing loads or drops in responsiveness at the same times. It can also correlate between any two SAN devices; this helps identify situations where two servers are overloading a switch, for instance.
When servers exceed their parameters, Performance Tuner sends alerts via e-mail or SNMP traps. Included filters set boundaries to send an alert only if usage exceeds a parameter a specified number of times or for a given duration. With this safety, a temporary spike in utilization won’t trigger a false alarm.
The SANView module can be run on its own or from within Storage Manager. It collects data on all the parts of the SAN, from host bus adapters and switches to storage devices and drives, and integrates with Storage Manager and Performance Tuner to provide user and diagnostics data for the SAN. It integrates nicely with Storage Manager and provides a vendor-agnostic view of the entire SAN, something few other vendors offer.
Both the Performance Tuner and SANView additional modules will be useful to most administrators, although Storage Manager’s Space Optimizer and Chargeback modules do a nice job on their own.
Any organization that needs to manage storage across multiple servers and storage platforms should consider Storage Manager 2.5.1. The reporting and diagnostic tools should pay off quickly by allowing you to identify peak users, find bottlenecks, and justify additional purchases through trend analysis. After all, optimizing a storage network can postpone the purchase of new hardware by reducing existing storage requirements.
Overall Score (100%)
|Storage Manager 2.5.1||9.0||9.0||9.0||9.0||9.0||9.0|
Having trouble installing and setting up Win10? You aren’t alone. Here are many of the most common...
Hot or not? From the web to the motherboard to the training ground, get the scoop on what's in and...
Confidence in our power over machines also makes us guilty of hoping to bend reality to our code
Microsoft says its new Azure cloud database is all types of databases in one. Here's why that might be...
Edge computing will not replace cloud computing, though the two approaches can complement each other ...
The Rust-like open source language tackles application development where asynchrony leads to...
The popular code repository is trying to be a one-stop shop for developers to get more of their work...