[ Mood: Cool ]
I was going through some of the pricing mechanisms of software in the Applications Management space. There are quite a few types of pricing. Well I have to say enterprises shell out out a lot for getting their software monitored.

Some of the pricing models I have come across are listed :

1) Pricing on a per monitored metric. Eg: Response time of webserver1

Here you need to spend based on the number of attributes of any server you need to monitor. Eg: you want to monitor response time and cpu usage of 2 systems, you need to pay for 2 x 2 = 4 metrics. You may be charged $xy / metric.

2) Some software charge you based on per server. You have an Oracle on a syetem and you want to monitor the oracle database, well you are charged say, $xyz / Oracle server (application).

3) Some also include the number of CPUs on the system in which the target application being monitored has. Hence if you have a webserver on a quad processor, you may be charged $xyz X 4 for monitoring just one instance of the software on this quad proc m/c.

4) Some are priced based on a server. Read that statement as per system. to be more clear, say you have an Oracle, WebLogic & webserver all installed on one Linux server. You pay per server. That’s it. You donot pay for per application on the Linux server.

5) Another software pricing model - monitor unlimited instances for a very low flat price. Well this looks very unconventional in the onset, but it looks like that’s what monitoring might be actually worth ? … maybe … Well probably it may not be worth pricing it on a monitor/server basis when the price is very low.

‘Monitor’ is a word that seems to be used differently. Some specify a monitor as a metric or attribute. Some refer an application as a monitor. Some confusion for users :-)