I have a requirement something like to get the detailed user list who are inactive from last 120 days means the users who are not made an attempt to login into an application from last 120 days. So could we get the same data by creating a BT..?
I believe this is not a trivial problem. Because to find inactive ID, we first need to have list of all UserIDs.
Also, I don't believe you can do temporal processing across multiple purepaths. It will require list or memory to be maintained to identify the inactive userids.
To achieve such kind of analysis you can exploit the Real Time BT feed in appmon to export the data. And from there have this analysed via full text search frameworks like ElasticSearch or Splunk.
Agree, we should have all the list of UserID's but could you please elaborate on how could we achieve this step by step after having the list of UserID's. And also can't we achieve this using Appmon alone.?
Thanks in Advance.
Idea is simple, Inactive IDs = Universe of Ids - Unique_List( UserIds in last 120 days).
Since User IDs will be unique, so you cannot store them within the PWH, however if there are less than 50,000 unique IDs (in splittings) you may give it a try. (Warning).
Then each day fetch this split report for 10-10 days. Then calculate the unique list.
Any Id not in this 120-day unique ID is theoretically inactive ID.
This is ofcourse a cumbersome hack. ;D And misuse of APM !