MQPerf standard service description

throughput optimums

MQPerf standard provides you with throughput and latency measurements related to the maximum sustainable throughput for JORAM on your own system according to 40 basic use cases of a JMS compliant MOM. The scenarios derive from 5 explored axes, with 2 or 3 possible values for each axis:

• destination type axis: using a JMS queue or topic,
• persistency axis: using persistent or transient messages,
• message size axis: using small (100 B/msg) or large (10 kB/msg) messages and your chosen size (from 0 to 100 kB/msg),
• connection type axis: the client connecting to JORAM through TCP (tcp, common JMS use) or being executed in the same JVM as JORAM (local, common J2EE use),
• parallel servers number axis: deploying 1, 2 or 4 JORAM servers on your system; this axis is explored with small messages only.

optimal configuration

The MQPerf standard report includes a description of the JORAM configurations used to get the reported optimum throughput. The configurations involve notably:

• JORAM persistency and transport modules,
• JORAM server parameters,
• JMS clients parameters.

The descriptions are also provided to JORAM users through J2EE (notably JOnAS) as configurations of the JORAM J2EE/JCA connector, and as parameters for the client beans.

behavioral analysis

MQPerf standard specifically analyses JORAM behaviour during 24 of the 40 preceeding tests. For each test the input load is set as a fixed profile based on the found optimum throughput. The load profile starts at the test optimum, then increases to a temporary overload, and concludes with a lower value. The system behaviour is analyzed with chosen indicators, whose values are collected all over the test and rendered into history charts. The indicators are:

• client indicators: production and consumption throughput curves displayed on the same chart;
• JORAM indicators: load index of the MOM execution engine, load index of the transactional persistency component, stickiness index of the messages flow related to the JORAM destination (queue or topic);

comparisons

The results of the tests on your system are compared to the results of the same tests on a number of reference platforms, including EC2 hosts types. Eventually your system is given a global performance score, which is then positioned in the scale of all the tests that have been previously submitted to the service.

Added value

MQPerf standard the JORAM project leader immediate solutions on some key aspects of his mission.

solving a performance problem

If the performance figures you get from an MQPerf community run answer your requirements, then you will find in the MQPerf standard report all the configurations and parameters to reach those performances in your project.

In addition the behaviour charts give you a better insight into the JORAM performances rationale, which may allow you to consider modifying your software architecture to better better handle the load.

See the more detailed use case.

responding to a call for tenders

The analysis report shows a high level of JORAM expertise. When you reuse elements of this report in your tender, you exhibit such expertise. The report is conveniently provided as a LibreOffice document, so that it is easy to reuse parts of it.

The actual execution of the tests also have a demonstration value, which makes your proposal all the more trustable. ScalAgent DT also allow you to use its name in your proposal, further enhancing the trust your readers will vest in the claimed results.

See the more detailed use case.

comparing with MQPerf community

MQPerf Standard does not replace MQPerf community. It provides additional services on key steps of the JORAM project leader activity. MQPerf community provides a preliminary diagnosis to the project leader.

for more details

You may appreciate in more details the value of MQPerf standard from the detailed report description here.