Armed and Capable

 
 
By Henry Baltazar  |  Posted 2003-03-24 Print this article Print
 
 
 
 
 
 
 


Armed and Capable

From a storage management standpoint, the frameworks ARM application is one of the most impressive plug-ins we saw in Control Center 5.1.1.

Storage management is no simple automation task, and this products agility in juggling all aspects of it—including the complexity of current SANs and important configuration variables such as zoning, LUN (logical unit number) masking and fabric management—is remarkable. Even more impressive is that ARM works with multiple hosts.

However, the ARM application currently works only with EMCs Symmetrix and Clariion arrays and with Hewlett-Packard Co.s StorageWorks arrays. We believe EMC must add support for other important RAID unit vendors to maximize ARMs market potential. EMC plans to support Hitachi Data Systems Corp. and IBM arrays in the future, officials said.

Once storage usage hits a preset threshold, ARM quickly carves out additional storage from an available array and presents it to the server. ARM automatically communicates with SAN switches (it currently supports switches from Brocade Communications Systems Inc. and McData Corp.) and ensures that servers have network links to the storage arrays.

Should any part of the automated procedure fail (including the RAID configuration or SAN switch configuration), ARM automatically discontinues operations and brings the SAN back to its previous working condition.

We were also very impressed with EMCs StorageScope application. StorageScope allows IT managers to create reports to see resource usage trends and more accurately plan future storage hardware acquisitions.

Like other SRM software weve seen, StorageScope did a good job of pooling data and presenting it in useful reports. Equipped with these reports, an IT manager could track down with relative ease the storage usage of specific users and departments and subsequently could set up chargeback services to justify IT spending.

Within the framework, Control Center 5.1.1 uses programmable response loops to keep things running as smoothly as possible. These loops determine who gets contacted in the event of a failure, such as an application fault, or if storage demand exceeds capacity.

In tests, we easily set up the response loop so that a system administrator would be the first person contacted in an emergency. We could also specify that an e-mail or page be sent to the backup administrator or an IT manager if the administrator didnt respond within a specified amount of time.

Control Center 5.1.1 can push SNMP traps out to other management tools, which is a good feature for organizations that have large investments in management packages such as HPs OpenView.

Control Center 5.1.1 also has some administrative granularity built in to it to help control multiuser access into the management console.

Control Center harnesses Windows 2000 to create and authenticate Control Center users; user and group policies can be created to restrict or expand access to specific EMC tools, such as StorageScope or RAID configuration tools.

We liked the fact that changes are tracked on a user-by-user basis because this is an important capability for auditing purposes.

Control Center 5.1.1 has some auto-discovery capabilities (to detect newly introduced storage units and databases), but dont expect it to discover everything on a SAN. It wont detect all SAN switches, for example, nor will it discover all database elements.

However, as Control Center continues to improve as a management suite, the auto-discovery functions should become more comprehensive.



 
 
 
 
 
 
 
 
 
 
 

Submit a Comment

Loading Comments...
 
Manage your Newsletters: Login   Register My Newsletters























 
 
 
 
 
 
 
 
 
 
 
Rocket Fuel