https://bz.apache.org/bugzilla/show_bug.cgi?id=59133
Bug ID: 59133
Summary: Regression in custom TG for upcoming 3.0
Product: JMeter
Version: unspecified
Hardware: PC
OS: Linux
Status: NEW
Severity: normal
Priority: P2
Component: Main
Assignee: [email protected]
Reporter: [email protected]
I've noticed important regression in current trunk code, which supposed to
become v.3.0. In BlazeMeter we have couple of plugins that are custom Thread
Groups. One of them, "Arrivals Thread Group" is supposed to operate with
"arrival rate" which is rate of "first sampler of thread group".
With JMeter 2.13 this thread group works as expected and produces valid load
level. With trunk code it produces exactly 1/2 of expected load. As additional
effect, we've noticed that threads that should be re-used for new arrivals are
finished and new threads are created.
For example test plan, there are total 120 arrivals supposed to happen. This
test plan is simpliest reproduction. Jmeter 2.13 gives expected 120 samples,
while trunk gives only 60.
I failed to find the source of this regression myself, I ask for help from
other contributors. I can't share the source code publicly due to legal
reasons, but can share it privately if this will help to solve regression.
--
You are receiving this mail because:
You are the assignee for the bug.
