Performance Modeling and Benchmarking of Event-Based ... - DVS
Performance Modeling and Benchmarking of Event-Based ... - DVS
Performance Modeling and Benchmarking of Event-Based ... - DVS
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Set Properties<br />
Set Properties<br />
Set Properties<br />
116 CHAPTER 5. BENCHMARKING OF EVENT-BASED SYSTEMS<br />
Scenario I<br />
OrderQueue 1<br />
Pub Sub<br />
… OrderQueue n<br />
Pub Sub<br />
Pub<br />
ConfQueue 1<br />
Sub<br />
… ConfQueue n<br />
Pub Sub<br />
Topics<br />
Other Queues<br />
SM 1 DC 1<br />
Recv<br />
SM n DC n<br />
Recv<br />
DC 1 SM 1<br />
Recv<br />
DC n SM n<br />
Recv<br />
Scenario II<br />
Other Topics<br />
Incoming Order Topic<br />
Order Confirmation Topic<br />
Pub<br />
Pub<br />
Sub<br />
Sub<br />
Pub<br />
Pub<br />
Sub<br />
Sub<br />
(DC=1)<br />
(DC=n)<br />
(DC=1)<br />
(DC=n)<br />
Selectors<br />
(SM=1)<br />
(SM=n)<br />
(SM=1)<br />
(SM=n)<br />
Selectors<br />
Recv<br />
Recv<br />
Recv<br />
Recv<br />
SM 1<br />
…<br />
SM n<br />
DC 1<br />
…<br />
DC n<br />
DC 1<br />
…<br />
DC n<br />
SM 1<br />
…<br />
SM n<br />
Scenario III<br />
Message Bus<br />
Pub<br />
Pub<br />
Pub<br />
Pub<br />
Sub<br />
Sub<br />
Sub<br />
Sub<br />
(DC=1)<br />
(T=Order)<br />
(DC=n)<br />
(T=Order)<br />
(SM=1)<br />
(T=Conf.)<br />
(SM=n)<br />
(T=Conf.)<br />
(DC=1)<br />
(T=Order)<br />
(DC=n)<br />
(T=Order)<br />
(SM=1)<br />
(T=Conf.)<br />
(SM=n)<br />
(T=Conf.)<br />
Selectors<br />
Recv<br />
Recv<br />
Recv<br />
SM 1<br />
…<br />
SM n<br />
DC 1<br />
… DC n DC 1 … DC n SM 1 … SM n<br />
Figure 5.32: Considered Scenarios<br />
• Scenario II (Pub/Sub with Multiple Topics): For each message type, a separate<br />
topic is used, i.e., the TD configuration parameter is set to MessageType (cf. Table 5.12).<br />
• Scenario III (Pub/Sub with Message Bus): One topic is used for all messages, i.e.,<br />
the TD configuration parameter is set to Central (cf. Table 5.12).<br />
The three scenarios differ mainly in terms <strong>of</strong> the flexibility they provide. While Scenario I is<br />
easy to implement given that no properties or selectors are necessary, it requires a reconfiguration<br />
<strong>of</strong> the MOM server for each new location or message type since new queues have to be set up.<br />
In contrast, Scenarios II <strong>and</strong> III, which only use topics, provide more flexibility. In Scenario II,<br />
a reconfiguration <strong>of</strong> the MOM server is necessary only when introducing new message types.<br />
Scenario III doesn’t require reconfiguration at all since a single topic (message bus) is used for<br />
communication. In addition, Scenarios II <strong>and</strong> III support one-to-many communication while the<br />
queue-based interactions in Scenario I are limited to one-to-one communication. One-to-many<br />
communication based on pub/sub allows to easily add additional message consumers, e.g., to<br />
maintain statistics about orders. On the other h<strong>and</strong>, the use <strong>of</strong> a limited number <strong>of</strong> topics in<br />
Scenarios II <strong>and</strong> III degrades the system scalability. As shown in the next section, the jms2009-<br />
PS benchmark allows to evaluate the trade-<strong>of</strong>fs that different configurations provide in terms <strong>of</strong><br />
flexibility, performance <strong>and</strong> scalability.