Roundtable Software Newsletter #32, March 18, 2005
In this issue...
|Patch 17: Credit Card Enhancement , Bank Rec for Multiple Banks & Inventory Printing|
|Patch 17 addresses the following:
Click Here To Download This Patch
|Unix Users: How To Get The Patch 17 Update|
|Patch 17 is also available (as Unix Patch 1) for your version 4 Unix product; to download it you'll need the login and password for the Unix Patches web page on our site. To get this information your 4.0 product must be registered. Send us an email stating the serial number of your 4.0 product and the contact name that was listed on your registration card and we'll send you the login and password. Once you have the login and password, you can go to the patches page and follow the link for Unix Patches at the top of that page, or click the link below.
Click Here To Download This Patch
|Q&A: Why are RG reports so much faster on a peer-to-peer server?|
|Q: I am on a peer-to-peer network and have noticed that when I print reports out of Report Generator they print much more quickly from the server than from the workstations. When I print similar reports from the other modules the speed difference between server and workstations is not nearly as pronounced. Why is that?
A: Before we discuss your specific question, a few words about peer-to-peer networks in general.
It is the nature of peer-to-peer networks that processing is much faster on the server than it is on a workstation. The reason is that the server has the data on its local hard drive, whereas the workstation must move all the data through a network connection. Even though networks are very fast these days, moving large amounts of data through a network connection is by its nature very slow by comparison.
Peer-to-peer networking is appealing because you save the price of one machine in your business. However, peer-to-peer networking is also very slow and inefficient in comparison to using a dedicated server. When considering the cheap price of computer hardware these days, using a peer-to-peer network in an environment where speed is an important issue can be penny wise but pound foolish. Dedicated servers are far more efficient at processing requests from workstations than non-dedicated servers, so that even though data must still be pushed through the bottleneck of a network connection, the speed sacrifice is not as great as it is in a peer-to-peer configuration.
However, getting back to your specific question, you are correct in noticing that the speed difference will sometimes seem magnified when using Report Generator reports as compared to other reports in the RTS-Advantage package. The root cause is still the speed of network throughput, but Report Generator, because of the way it does things, really accentuates the difference between the server’s direct access to the data as compared to the workstation’s remote access.
The basic difference between Report Generator and other printing programs in RTS-Advantage is that RG is written to work with any data file in the RTS-Advantage system, while the other print programs are written to deal with certain specific files and work with them in specific ways. The printing programs in RTS-Advantage are each optimized to take best advantage of the specifics of their jobs, while Report Generator sacrifices optimization for generalization. Think of Report Generator as Cal Ripken, Jr, a generalist who could do everything well on the ball field, as compared to Mark McGwire, a specialist in just one thing, the home run. We promise, though, that RTS-Advantage is steroid-free!
The concept can be made more understandable with an example. Let’s say that you want to print a simple inventory item list. You want a list of all your items with Group Numbers starting with A. Your Report Generator report is going to look through all the items in your Inventory looking for and printing those with the qualified group number. If you happen to have 20,000 item numbers, that means RG will have to look at quite a lot of data to find those records. Every one of those 20,000 records will have to make the trip through the network pipeline to your workstation. On the other hand, when you print the list out of Inventory program 1, this program knows that there is a key defined on the Group Number field that can be used as a shortcut to read only those records that actually qualify for the report. So if 100 records have a qualifying group number, this report reads only 100 compared to Report Generator which reads 20,000! This means that Report Generator has to read 200 times more data than the corresponding Inventory program. As you can imagine, sending that much more data across a network connection can make for a noticeable speed difference.
Report Generator cannot take advantage of all the shortcuts and special knowledge used by the other print programs in RTS-Advantage; if it tried to RG would be so bloated and unwieldy as to be practically impossible to maintain. And on a single-user or dedicated network system where data throughput is not a major issue the difference is usually inconsequential.
Due to the nature of peer-to-peer networking, any data-intensive task--be it in RTS-Advantage or any other application--will be slower when run across the network than when run on the server. The print programs in RTS-Advantage have been optimized to process as fast as possible no matter the networking environment. Report Generator, being a general print engine, can only be optimized so far and therefore reports printed through RG can take noticeably longer to process when run from a workstation as compared to being run from the server. If performance is important to you, consider a dedicated server.
|Roundtable Software 30831 Cove Road Tavares, FL 32778-5164
(352) 253-9779 FAX: (815) 572-5446
All contents copyright © 2000-2008 Roundtable Software. All rights reserved.