What count are you looking at? Regardless, I believe the answer will be that the failed ones are included in any counts. For instance in a business transaction if you chart successful count and failed count together and 100% are failing, you'll see two bars (in my case) of equal height.
Thanks James for your response,
Ok now, how to separate the two? This is because, my Bosses want to report this well and should be shown in both web and client dashboards clearly. Kindly help
I don't know of a way to directly achieve that. In theory you could make a BT that filters out any PurePaths with exceptions, error logging, and 'bad' http codes but then you would need a second BT to track the failed ones but this doesn't seem very practical.
Working with what you have you can chart failed and total in one area type chart so that the failed series is in front of the overall. This would give a good visualization of the number that are failing compared to the total count. You can also chart the percentage failed to get an idea of this.
In the longer term this may be a good candidate for an RFE/idea.
".....In theory you could make a BT that filters out any PurePaths with exceptions, error logging, and 'bad' http codes....."
How could you achieve that?? At least with that i can determine the successful and failed ones.
There are measures for the count of exceptions, logging (which you may need to refine to look at only certain levels of logging in the details of the measure), and HTTP status codes which would need to be set to look at 'bad' codes (e.g. 4xx, 5xx). You would need to create new measures for each of these with the lower severe threshold set to zero and then use these as filter measures in the business transaction. The idea would be to filter out any purepaths that have any such of those that would mark the purepath as failed. If this works and is configured correctly the failure rate would be zero.
That being said, personally I would find such a count to be of limited value as that failed count and more importantly the failure percentage relative to the overall count are what would indicate issues.
Thanks James very much.
Just one more thing to get clarified of,after having filtered the successful purepath/transactions, i think there is no need to create measures for the failed ones because at least those ones are created by default, right??