The University of Pennsylvania's Online Computing Magazine
|
 |
March 1993 - Volume 9:5 [Printout | Contents | Search ]
By Dr. Michael W. Kearney Imagine the following situation: You arrive at your lab early one morning to discover that a water hose has failed and drenched your desktop workstation. It's a total loss. To your credit, you have learned (after considerable effort) to operate your system's backup utility and have been using it to back up your hard disk regularly. Unfortunately, the backup tapes were stored right next to the workstation and they're waterlogged and unuseable too. What do you do now? Although the details and severity will differ from this example, the fact is that some day you are going to lose all or part of the data stored on your desktop system. Simple prudence suggests that you should take steps to avoid such losses. Unfortunately, doing this job right often demands an increased investment in hardware, software, storage media, and, most significantly, your time. You must master the operation of your system's backup utility (often completely different from one system to the next), run it regularly, manage numerous diskette or tape volumes, and try to find the right files to restore when losses occur. This can be a time-consuming and error-prone activity. Often, the result is that desktop systems are not backed up as frequently or as thoroughly as they ought to be. PENNback, a new service from Information Systems and Computing and the University Data Center (UDC) solves many of these problems. With it, you can back up your desktop system easily and reliably. PENNback saves you time by eliminating much of the day-to-day drudgery required for routine backup. Both you and your organization benefit since you have more time to devote to other challenges of your job.
How does PENNback work?PENNback currently uses client/server software from IBM called the Workstation Data Save Facility/VM. (See Penn Printout, February 1992, page 1, for an explanation of client/server computing.) For the purpose of this discussion, you need only understand that the PENNback software comes in two parts. The first part, called the client, runs on your desktop system. The second, called the server, currently runs on the University Data Center's IBM 9121 processor.PENNback's client software backs up your disks by sending a copy of each disk file to the server over a computer network, usually via a PennNet Ethernet connection. The PENNback server software, in turn, stores each file copy it receives on UDC-managed disks and tapes. At the same time, it maintains a database containing a complete record of your backup activity. In the event you need to restore a file, an easy- to-use interface presents a list of backed up files on your desktop display. You select the file(s) you want to restore from the list and press a key. The PENNback server extracts the location of each file copy from its database, reads it from wherever it's currently stored (usually tape), and returns it to your system via the network. That's all there is to it.
PENNback highlightsPENNback supports many popular desktop systems. Here is the current list:
PENNback is fully supported. Training, consulting, client software, documentation, problem resolution, and server operations are included with PENNback services. PENNback does not require you to manage backup media. Backup data is kept in a centrally managed repository. This reduces significantly or eliminates many of the headaches involved in managing, organizing, and handling tape or diskette backup volumes. PENNback is easy to use. PENNback clients are easy to install and configure and provide an easy-to-use interface for file backup and restore. PENNback provides off-site storage of backup data. The backup data is physically separated from the original copy, reducing the likelihood that a single event could cause the loss of both the data and its backup, as in our watery example above. PENNback provides file archiving. PENNback can archive important but infrequently used files, allowing you to erase your local copy. This can save you considerable disk space if the archived files are large. PENNback uses standard communications and protocols. PENNback clients and servers communicate using TCP/IP, a world-wide and campus standard. In many cases, if your system already has supported TCP/IP software and a PennNet Ethernet connection, the only initial investment you need to make is to acquire a copy of the client software.
Does PENNback work?PENNback has been up and running since July 1, 1992. From October through December, UDC ran a comprehensive pilot study for a selected group of 20 systems. All supported configurations were tested under realistic conditions and more than 350,000 customer files containing over 10 gigabytes of data were backed up with with no reported losses of files or data. If you'd like to talk with a current PENNback customer, UDC will be happy to provide references.
What does PENNback cost?A client installation pack costs $100. It includes a copy of the software, future updates, a manual, and training for installation and use. If you are responsible for multiple systems, you need buy only one copy of the client software for each system type you're supporting.Charges start with a $10 per month base fee. Additional charges depend on the number of files, amount of data, and frequency of use. Total monthly charges (base charge plus additional) for single-user systems, such as Macintosh or IBM PC, typically amount to about $20. For a large server configuration (2 to 3 gigabytes of data, 150,000 files) the charges are about $70 per month. Of course, your "mileage" may vary. UDC will give you an estimate for your approximate configuration and intended use.
Getting started with PENNbackTo get started using PENNback you must:
DR. MICHAEL W. KEARNEY is Associate Director for Academic Services, University Data Center.
|