Take the 2-minute tour ×
Ask Patents is a question and answer site for people interested in improving and participating in the patent system. It's 100% free, no registration required.

Relates to Patent Application 20090055833 - "System and Method for performance monitoring."

I come with two questions; First, it would seem to me substantial prior art exists to 1994. However, this patent obfuscates and applies supposedly proprietary methods heavily to attempt to create a new patent able process out of software elements. To wit; the alleged patent details a methodology for logging relevant performance data to a specified database and mechanical interface. It's really a very cut and dry and buttered patent to my eyes, but that's all things IBM..

Question two, to wit, the first blush of this tells me that this is not a new or novel patent - simply a formalization of something which was commonplace for any sort of performance monitoring (host-based or agent-based) for the era. To the extent that this patent is painfully overbroad in a dangerous fashion: Claim 1: A method for monitoring a computer software system by reading log records written by said software system to determine performance of said software system relative to response time to end users, comprising: adjustably tuning performance evaluation bias by a computer software monitoring system between processor and memory consumption; responsive to said bias, monitoring performance of said computer software system with respect to transaction time parameters including said response time to end users; and receiving from a user a first tuning parameter for allocating memory for said monitoring performance and a second tuning parameter for specifying a time out value for in-flight units of work; wherein a in-flight unit of work is timed-out if a in-transit time for the in-flight unit of work exceeds the time out value to prevent a single unit of work from having a high impact on the monitored transaction time parameters. LAYMAN: By reading logs, we apply math to come up with a performance number.

  1. A system for monitoring a computer software system by reading log records written by said software system to determine performance of said software system relative to response time to end users, comprising: a non-transitory computer usable medium having computer readable program code logic comprising: a first user actuated tuning knob for allocating space in memory for performance monitoring; a second user actuated tuning knob for a specifying a time out value for in-flight units of work; and a transaction monitor responsive to said first and second user actuated tuning knobs for accumulating one or more transaction time parameters for a plurality of said in-flight units of work, wherein a in-flight unit of work is timed-out if a in-transit time for the in-flight unit of work exceeds the time out value to prevent a single unit of work from having a high impact on the transaction time parameters. LAYMAN: This says plainly that data storage and even processing is handled on an external server to the one being actively monitored. One which has no purpose but to analyze this specific data and present results to user.
  2. A program storage device readable by a machine, tangibly embodying a program of instructions executable by a machine to perform method steps for monitoring a computer software system by reading log records written by said software system to determine performance of said software system relative to response time to end users, said method comprising: adjustably tuning performance evaluation bias between processor and memory consumption; responsive to said bias, monitoring performance of said computer software system with respect to transaction time parameters; and receiving from a user a first tuning parameter for allocating memory for said monitoring performance and a second tuning parameter for specifying a time out value for in-flight units of work; wherein a in-flight unit of work is timed-out if a in-transit time for the in-flight unit of work exceeds the time out value to prevent a single unit of work from having a high impact on the monitored transaction time parameters. LAYMAN: This would seem to describe yet another independent computer which does nothing but parse these logs, these inputs, and then provide to the end user or manager numbers derived through this method.

Basically, it looks like a cobbled together performance monitor for the purposes of scheduler operations. I'm not good at reading patents, even after taking classes, so maybe I'm missing something here? This seems extremely broad - to the extent where anyone doing any sort of performance monitoring using a similar system (virtually all of them) is in violation.

Is Prior Art available, or another similar sample which was determined as overbroad?

share|improve this question
add comment

1 Answer

One thing missed is that the specifc document you point to is not a patent at all. It is an application for a patent. This application did result in three patents including US8181160. As mentioned in other answers at Ask Patents, PAIR has an unforgiving user interface. Look at the examples it shows for number formats and use the radio buttons to tell it what kind of a number you are giving it. The "application number" 2009/123456 is the publication number of the application, not the application number.

You can look up its history in USPTO Public PAIR.The entire back-and-forth between the applicant and the examiner are there as well as all the things the applicant and the examiner knew of as potentially relevant prior art are listed. A quick look at the image file wrapper shows the claims were rejected and correspondingly narrowed/clarified three times between the claims in the initial application and the granted claims.

One of the basics of patent reading are that the finally granted claims define what the examiner agreed was new.

share|improve this answer
    
Ah-ha. This one got passed on to me and I remembered the site. The problem I had is that apparently, I'm bad at operating PAIR. (Plugging in the application just said 'not found.') So that problem solved, looking at US8181160, am I reading this wrong or did they patent memory profiling? –  RootWyrm Aug 1 '13 at 20:52
    
In the very early days of computing IBM did patent things like interrupts and subroutines. The patent you have linked to is considerably more abstract - i think it would take a half an hour to try to figure out what it really thinks it covers. (Btw I do not think your link points to the patent US8181160, it took me to 6,817,011) –  George White Aug 2 '13 at 0:16
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.