The Prevail Case Management System: Technical Support Articles

The Machine That Goes PING!

No, I'm not referring to the infamous hospital machine in the Monty Python sketch from "The Meaning of Life." I'm referring to something considerably less entertaining, but infinitely more important to the health and welfare of your practice: Your network.

Nearly a third of the support calls we receive about "Prevail" problems have nothing to do with Prevail. They're strictly network problems. Prevail just happens to be the first application that transfers enough data across the network at one time for you to notice the problem. Regardless of its type or sophistication, your network is just an infrastructure set up to send data back and forth between computers. You can think of it as a garden hose connected to a faucet at one end and a new-fangled sprinkler gizmo at the other end. Opening up a Word or WordPerfect document is the equivalent of opening the faucet maybe a 1/10 of a turn— a very small amount of data flows through the "hose". If you should happen to have a couple of holes in the hose or the washer needs to be replaced, nobody notices because the pressure isn't sufficient to produce anything more than a drip— that is, you're not sending very much data across the network.

Now, let's say you load Prevail. This is the equivalent of turning the faucet on full-blast. Every pinhole in the hose suddenly becomes a mini-geyser, and the bad washer at the faucet end of those sprays water all over your pants before you can get the damn thing turned off.

In the sprinkler world, most people would just fix the leaks and get on with their lives... but not in the computer world. In the computer world, the first thing they do is pick up the phone and call the sprinkler gizmo people (namely us) to find out why it's not spinning around as fast as it should.

Maybe it's just because you can't see the problem in a network like you would with a garden hose... but it's there nonetheless. If you do have a network problem, Prevail will let you know about it. You'll very likely notice one or more errors indicating that your network needs some attention. Here are just a few:

Prevail Cannot Find Network Directory - This error usually occurs when starting Prevail if the server drive is not mapped correctly, or if the workstation is not on the network.
7009 Problem with Advantage server file write - This error usually occurs when there is not enough free disk space on the server, or the workstation is not connected to the network.
7010 Problem with Advantage server file read - This error usually indicates some form of data corruption usually caused by network errors.
7011 Problem with Advantage server file seek - The specified record cannot be accessed usually because the record is corrupt.
7012 Problem with Advantage server file offset - The specified record cannot be accessed because of a corrupted index.
7038 Invalid record number - This error indicates index corruption which can be caused by network failures.

Resolving network problems can involve any element in the system: Network cards, network card drivers, network hub, server, etc. This is normally something that requires the services of a good network technician. If you don't have one handy, however, there are a few simple tests you can run on your own to help isolate the problem. That's where we get to the PING part!

From a MS DOS prompt (Start/Run/Programs/MS-DOS Prompt), you can use the PING command to send data to other machines and listen for the return of those signals, just like submarines do with sonar. To do this, you'll need to know either the network computer names of the other machines on your network (which you can see in your Network Neighborhood).

The syntax for using the PING command is follows:

PING -N 100 -L 20000

For example, if your server is called SERVER on the network, you would type

PING SERVER -N 100 -L 20000

This would ping the server computer with 20,000 bytes of data 100 times and display the results to the screen (this will take a few minutes). The most important thing to look for in the reported results is the percentage of packets that are lost. If everything is in order, the packet loss percentage will probably be zero. Even if it's not, you're not out of the woods yet. You'll need to ping test every machine on the network at various times throughout the day so you'll get results under differing network traffic loads. Sometimes a machine with a bad network card will test perfectly one minute and return all sorts of errors the next. The difference is generally the amount of other network traffic going on at the time of the test.

If you get errors, timeouts or high packet loss, you've probably got a problem... but where? If all machines get errors, the problem is likely to be the network hub, cabling, or the server itself. If one machine gets errors and the others don't, the problem is more likely to be with the network card and/or cable connection with that computer itself.

Unless you're a real tech-head, fixing the problem is going to require the services of the aforementioned network technician. That's high-level stuff which falls beyond the scope of this article... but at least you'll be able to give your network guy some meaningful information when he shows up.

That's a big step in the right direction.