I've been working with RPT 126.96.36.199 and now 188.8.131.52 primarily with web service tests. I've been experiencing extremely poor performance when recording / editing service tests with larger (i.e.: > 100 KB) responses.
For example, a 500 KB response is causing the recording UI to become completely unresponsive for over 20 minutes from the time the response is received (i.e.: the service request has been processed and the response "packet" received by RPT to parse). During this time the RPT Java process rapidly consumes more RAM, reaching over 1 GB allocated for a response of this size, and consumes 100% of a core on a 4 core processor. Larger responses consume more RAM and take longer to process: upwards of an hour and 4+ GB of RAM for 15 MB responses.
I am running RPT 184.108.40.206, configured with a maximum Java heap of 8192 MB (-Xmx8192m) on a machine with 16 GB of RAM. No swapping/virtual memory is being used during processing. The OS is RedHat Linux 6.4 with current patches. If I'm very patient, the test (request and response) is successfully recorded / edited. But as you can imagine, this makes test development rather frustrating.
Note that this is not related to executing the test: my query only pertains to recording and editing. Is this "normal" behaviour for RPT 8.5 and later while recording web service tests? Are there any alternatives or fixes known?