Transcations Exceeding threshold
Hi,
I have a URI abc.xhtml whose upper warming threshold is set to 5 seconds.I have a created a measure for this URI with 5 seconds threshold.
Can you please help me with the following:
a) How can I prepare a dashlet that shows me only the transactions for this URI that exceed the threshold within the selected time frame.
b) How can I get the count of the transactions that have exceeded the threshold within the selected time frame.
The incident will fire when there is a total count of 10 in the past 30 minutes. It will continue to evaluate that count, every 10 seconds, and the incident will stay active as long as there is a count of 10 over the past 30 minutes. From your example I can't see exactly 21;41, but what I do see is that in the period 21:43 back to 21:13 there is a count of (1+1+2+4=8) violations, which is less than 10 so it would stop.
Hi Andreas.My aggregation value is Count.This is inline with your suggestion in one of your email to me.. You has sent me an email with the screenshots few days back
Hi,
I am able to get the incidents but I am not able to depict how dynaTrace is calculating these values
| Count of Pages exceeding threshold | Value |
| 2014-04-29T20:38:00 | 1 |
| 2014-04-29T20:39:00 | 1 |
| 2014-04-29T20:40:00 | 3 |
| 2014-04-29T20:41:00 | 1 |
| 2014-04-29T20:42:00 | 2 |
| 2014-04-29T20:43:00 | 4 |
| 2014-04-29T20:45:00 | 10 |
| 2014-04-29T20:46:00 | 4 |
| 2014-04-29T20:47:00 | 1 |
| 2014-04-29T21:08:00 | 1 |
| 2014-04-29T21:09:00 | 4 |
| 2014-04-29T21:10:00 | 3 |
| 2014-04-29T21:11:00 | 4 |
| 2014-04-29T21:12:00 | 2 |
| 2014-04-29T21:13:00 | 4 |
| 2014-04-29T21:39:00 | 2 |
| 2014-04-29T21:40:00 | 1 |
| 2014-04-29T21:43:00 | 1 |
I have set an incident rule to throw up an alert when the count reaches 10 within a time frame of half an hour. My alert details are as follows:
| Start Time | Duration | End Time |
| 2014-04-29T20:43:30 | 58 Mins 10 Seconds | 2014-04-29T21:41:40 |
The alert has triggered at the correct time but I am not able to figure out why did it last for 58mins and end at 21:41. My evaluation timeframe is et to 30 mins only.
Whats your "Aggregation" Setting in your Condition. If you e.g: have "Max" then the incident stays up as long as the Max Value in the last 30 minutes is above your threshold. If you e.g: specify Average then it stay up as long as the Average value of the last 30 minutes is above the threshold. So - check your Aggregation setting
I believe the way in which we have set the rule is taking the failed transaction count (like http response failures etc ) instead of the count of transaction exceeding threshold. Can you advise on how to get this corrected ?
My business transaction is correctly listing down the transactions exceeding 10 second threshold .We have used this BT in the Incident rule.
Hi Andreas,
It will be difficult to export the data from Prod. as we have restricted access. Will the screenshots of the pure path help or do we definitely need a PP export.
PFA, the screenshot of the Incident Dashboard showing no alerts . I hope the incident rule is not taking into account the transaction failed count ( for eg: http response errors, exceptions etc) .
Can you export a couple of PPs and send it over? Either attach it to this forum or send it to me via email
Hi Andreas,
I have setup the incident rule as suggested in the above screenshots but I am not getting any alerts out of it.
Basically I had the breaches happening as shown in the below screenshots but alerts are not coming up.Please help !!
Can you check your incident dashboard to verify that you really dont get this incident? Based on the screenshot it seems that everything is correct
Thanks Andreas !! I have set the incident rule accordingly.
As Richard suggested, I created measures for "Web Request Response time" for 5-6 URL's and added the threshold to them(say 5 seconds). Then I added all these measures to the same Business Transaction( as specified in trail above). This business transaction is able to give me the correct count of transactions specific to these URL's exceeding the specified threshold.
a) In the real world scenario, I have some 200 URL's whose SLA needs to be measured for the same threshold. Is there any other effective way to handle this scenario or do I need to create a measure for all the 200 odd URL's and then add all of them to the same business transaction to get the desired output?
b) Will it not have any performance impact /
You need to select the "Failed Transaction Count" Result Measure of your Business Transaction. If you look at your "Select Measure" Screenshot from yesterday that shows your Business Transaction you see that there is a "Failed Transaction Count" Result Measure. If you use that for your Condition and then Edit that measure you can specify the Count Threshold. This is exactly what I did in my screenshot
Hi Andreas,
I have a Business Transaction that shows the count of transactions that have exceeded the threshold.This BT is internally calling many measures (for different web pages) as specified in the Richard's email below.The problem that I am facing is when I am trying to create an incident for this transaction and trying to set the threshold ,its giving not giving me the option to set the threshold on count ,instead its giving the option to set the threshold on Response time.
