In a very frustrating and longer than necessary debug session with a LRC engineer, we were UNABLE to associate the
VUSER ID which is listed under the VUSER ID column of the ERROR.CSV file to the VUSER_XX.LOG file that contains
the detailed logging information surrounding that error. For example, if the VUSER ID under the VUSER ID column was 1
you would expect that VUSER_1.LOG would be the log file that contains alll logging information associated with a VUSER
ID of 1. But that is not the case.
As a result, we have to search all 210 VUSER_XX LOG files . What was even more frustrating is that with the MAXIMUM
size limit of 1024 MB or 1 GIGABYTE for all Vuser_XX log files, many of the 210 log files had no information at all since the
MAXIMUM limit of 1024 MB had been exceeded only after running the load test for about 10 min.
Besides the requirement to be able to easily find the correct VUSER_XX.LOG file quickly, there needs to be way for the
LRC engineer to TEMPORARILY INCREASE the MAXIMUM size limit.
In my case, Loadrunner R&D required the full detailed logs, but for many performance-related problems it takes time to
generate the performance issue in an web application. Consequently, the need for more flexibility in assigned the
MAXIMUM size is sorely needed. Besides, DASD space for storage is one of the cheaper resources in a CLOUD
environmemnt when it is compared to additional RAM requirements or additional CPU requirements.
In my case dealing with only 210 Vuser, I am only using less than 10% of the total Vuser Web-HTML capability of a
SINGLE Cloud-based load generator. Nonetheless, the FULL logging request from Loadrunner R&D is causing me to
run into the ROADBLOCK of inadequate logging space that such a small load test that I should not have to deal with.