1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
|
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML//EN">
<html>
<head>
<title>Real-time Event Channel Configuration</title>
<!-- $Id$ -->
</head>
<body>
<CENTER>
<h1>Real-time Event Channel Configuration</h1>
</CENTER>
<H3>Overview</H3>
<P>The new implementation of the real-time event channel uses a
factory to build all the objects and strategies it requires.
The factory can be dynamically loaded using ACE Service
Configurator, this is extremely convenient because the factory
can also parse options in the Service Configurator script file.
</P>
<P>The current implementation provides a default implementation
for this Factory, this document describes the options used by
this default implementation.
</P>
<H3>The configuration file</H3>
<P>The real-time event channel uses the same service configurator
file that the ORB uses, the default name for this file is
<CODE>svc.conf</CODE>, but the ORB option
<CODE>-ORBSvcConf</CODE> can be used to override this.
The format of the file is described in detail in
the service configurator documentation, but the relevant section
for the event channel looks like this:
</P>
<PRE>
# Comments go here...
# More comments if you want to...
static EC_Factory "-ECFiltering basic ....."
</PRE>
<P>All the event service factory options start with
<CODE>-EC</CODE></P>
<H3>The options</H3>
<P><TABLE BORDER="2" CELLSPACING="2" CELLPADDING="0">
<TR>
<TH>Option</TH>
<TH>Description</TH>
</TR>
<!-- <TR NAME="ECDispatching"> -->
<TR>
<TD><CODE>-ECDispatching</CODE>
<EM>dispatching_strategy</EM>
</TD>
<TD>Select the dispatching strategy used by the real-time
event service. A <EM>reactive</EM> strategy will use the
same thread that received the event from the supplier to
push the event to all the consumers,
the <EM>priority</EM> strategy will use a prioritized pool
of threads and will query the scheduling service to select
the thread that will dispatch each event;
the <EM>mt</EM> strategy will also use a pool of threads,
but the thread to dispatch is randomly selected.
</TD>
</TR>
<!-- <TR NAME="ECDispatchingThreads"> -->
<TR>
<TD><CODE>-ECDispatchingThreads</CODE>
<EM>number_of_threads</EM>
</TD>
<TD>Select the number of threads used bythe <EM>mt</EM>
dispatching strategy.
</TD>
</TR>
<!-- <TR NAME="ECFiltering"> -->
<TR>
<TD><CODE>-ECFiltering</CODE>
<EM>consumer_filtering_strategy</EM>
</TD>
<TD>Select the filtering strategy used by the consumers.
The <EM>null</EM> filtering strategy will build trivial
filters for all consumers.
The <EM>basic</EM> filtering strategy supports
disjunction, conjunctions and timeouts based on the
subscriptions passed by the consumer during the connect
call.
The <EM>priority</EM> filtering strategy supports the same
features, but it also collaborates with the scheduling
service to build the dependency graph.
</TD>
</TR>
<!-- <TR NAME="ECSupplierFiltering"> -->
<TR>
<TD><CODE>-ECSupplierFiltering</CODE>
<EM>supplier_filtering_strategy</EM>
</TD>
<TD>The event channel can also perform some filtering close
to the suppliers by minimizing the number of consumers
that are tested for given event.
If the strategy is <EM>null</EM> then a global collection
of consumers is maintained and each event is filtered by
each consumer.
If the strategy is <EM>per-supplier</EM> then the EC uses
the supplier publications and the consumer subscriptions
to determine which consumers could potentially be
interested in events for each supplier.
This set of consumers is usually smaller than the complete
set and it is thus faster to traverse it, but keeping more
collections of consumers increases the connection and
disconnection time as well as the memory requirements.
</TD>
</TR>
<!-- <TR NAME="ECTimeout"> -->
<TR>
<TD><CODE>-ECTimeout</CODE>
<EM>timeout_strategy</EM>
</TD>
<TD>A consumer can request that the event channel generate
periodic or oneshot timeout events.
This option controls the strategy to generate the
timeouts,
using <EM>reactive</EM> the same reactor used for the ORB
is used by the event service.
The <EM>priority</EM> strategy uses a prioritized group of
threads, timeouts with higher rate are generated by
threads of higher priority.
<BR><B>NOTE: The <EM>priority</EM> strategy is not
implemented</B>
</TD>
</TR>
<!-- <TR NAME="ECObserver"> -->
<TR>
<TD><CODE>-ECObserver</CODE>
<EM>observer</EM>
</TD>
<TD>The event channel uses the Observer pattern to report
changes in the subscriptions and publications of its
suppliers and consumers; this is specially useful in the
implementation of event channel gateways.
The user can disable this feature by selecting the
<EM>null</EM> strategy;
whereas the <EM>basic</EM> strategy provides a simple, yet
efficient implementation.
</TD>
</TR>
<!-- <TR NAME="ECScheduling"> -->
<TR>
<TD><CODE>-ECScheduling</CODE>
<EM>scheduling_strategy</EM>
</TD>
<TD>The event channel can collaborate with the scheduling
service to build the dependency list between the consumers
and the suppliers.
If the <EM>null</EM> scheduling strategy is selected this
feature is disabled,
the <EM>priority</EM> strategy enables this feature.
<BR><B>NOTE:<B> The default is to have the feature disabled.
</TD>
</TR>
<!-- <TR NAME="ECPushSupplierSet"> -->
<TR>
<TD><CODE>-ECPushSupplierSet</CODE>
<EM>strategy</EM>
</TD>
<TD><P>Consumers can connect and disconnect from the event
channel as part of push to one consumer;
changing the set of consumers as a result of such an
operation would invalidate the iterators used by the event
channel to dispatch to the relevant consumers.
There are several ways to handle this problem, for example
a copy of the set could be made <I>before</I> initiating
the dispatch, or only if there is a change on the set
<I>during</I> the dispatch.
If the dispatching strategy is not reactive then it is
possible to lock the set of consumers during the duration
of the dispatch operation, the change in the subscriptions
will be held until the event is completely processed.
Finally the subscription change could be delayed until the
event is fully processed and there is no risk of
invalidating an iterator.
</P>
<P>All this strategies have merits under different
circumstances, the user can select the right one for his
application using this option.
The <EM>immediate</EM> strategy will perform any changes
on the consumer set immediately, simply using a lock to
synchronize access. The <EM>delayed</EM> operation will
wait until there are no threads iterating over the set
to perform any modifications on it.
The <EM>immediate_st</EM> simply performs the
modications without taking any locks, it is useful for
single threaded applications that do not receive changes
are part of an upcall (for example: if there are no
collocated consumers).
</P>
</TD>
</TR>
<!-- <TR NAME="ECProxyConsumerLock"> -->
<TR>
<TD><CODE>-ECProxyConsumerLock</CODE>
<EM>lock_type</EM>
</TD>
<TD>Select the lock type (<EM>null</EM>, <EM>thread</EM> or
<EM>recursive</EM>) to synchronize access to the
ProxyPushConsumer state.
</TD>
</TR>
<!-- <TR NAME="ECProxySupplierLock"> -->
<TR>
<TD><CODE>-ECProxySupplierLock</CODE>
<EM>lock_type</EM>
</TD>
<TD>Select the lock type (<EM>null</EM>, <EM>thread</EM> or
<EM>recursive</EM>) to synchronize access to the
ProxyPushSupplier state.
</TD>
</TR>
<!-- <TR NAME="ECConsumerAdminLock"> -->
<TR>
<TD><CODE>-ECConsumerAdminLock</CODE>
<EM>lock_type</EM>
</TD>
<TD>Select the lock type (<EM>null</EM>, <EM>thread</EM> or
<EM>recursive</EM>) to synchronize access to the
ConsumerAdmin state.
</TD>
</TR>
<!-- <TR NAME="ECSupplierAdminLock"> -->
<TR>
<TD><CODE>-ECSupplierAdminLock</CODE>
<EM>lock_type</EM>
</TD>
<TD>Select the lock type (<EM>null</EM>, <EM>thread</EM> or
<EM>recursive</EM>) to synchronize access to the
SupplierAdmin state.
</TD>
</TR>
</TABLE>
</P>
<H3>The constructor</H3>
<P>The <CODE>TAO_EC_Event_Channel</CODE> class implements the
<CODE>RtecEventChannelAdmin::EventChannel</CODE> interface;
this class takes one mandatory and two optional parameters in
its constructor:
</P>
<PRE>
TAO_EC_Event_Channel (const TAO_EC_Event_Channel_Attributes& attributes,
TAO_EC_Factory* factory = 0,
int own_factory = 0);
</PRE>
<P>The <CODE>factory</CODE> is an optional parameter to override
the default strategy factory used by the event channel,
the event channel will destroy the factory if the
<CODE>own_factory</CODE> argument is true.
</P>
<P>The <CODE>attributes</CODE> parameter can be used to fine tune
some of the algorithms and strategies used by the event channel,
the default values are probably OK for most applications.
Notice that the attributes include the POA used to activate the
ConsumerAdmin, SupplierAdmin, ProxyPushConsumer and
ProxyPushSupplier objects;
this POAs must have the <CODE>IMPLICIT_ACTIVATION</CODE> and the
<CODE>SYSTEM_ID</CODE> policies (as the RootPOA does).
</P>
<P><TABLE BORDER="2" CELLSPACING="2" CELLPADDING="0">
<TR><TH>Attribute</TH><TH>Description</TH></TR>
<TR><TD><CODE>consumer_reconnecto</CODE></TD>
<TD>If the attribute is not zero then the same consumer can
call <CODE>connect_push_consumer</CODE> on its
ProxyPushSupplier multiple times to change its
subscriptions;
this is usually more efficient that disconnecting and
connecting again.
</TD>
</TR>
<TR><TD><CODE>supplier_reconnecto</CODE></TD>
<TD>If the attribute is not zero then the same supplier can
call <CODE>connect_push_supplier</CODE> on its
ProxyPushConsumer multiple times to change its
publications;
this is usually more efficient that disconnecting and
connecting again.
</TD>
</TR>
<TR><TD><CODE>busy_hwm</CODE></TD>
<TD>When using the delayed strategy to update
ProxyPushSupplier sets this flag controls the maximum
number of thread that can simultaneously iterate over the
set before blocking.
It can be used to avoid starvation in delayed updates on
the set.
</TD>
</TR>
<TR><TD><CODE>max_write_delay</CODE></TD>
<TD>When using the delayed strategy to update
ProxyPushSupplier sets this flag controls the maximum
number of threads that will initiate dispatching
<B>after</B> a change has been posted.
Any thread after that is blocked until the operations are
performed.
It can be used to completely stop starvation of delayed
updates on the set.
</TD>
</TR>
<TR><TD><CODE>scheduler</CODE></TD>
<TD>Most configurations of the real-time event channel do
not require an scheduler.
If any of the strategies that require an scheduling
service is selected then this attribute should be set
appropiately.
</TD>
</TR>
</TABLE>
</P>
<P><HR><P>
Back to the TAO <A HREF="index.html">components documentation</A>.
<!--#include virtual="/~schmidt/cgi-sig.html" -->
<address><a href="mailto:coryan@cs.wustl.edu">Carlos O'Ryan</a></address>
<!-- Created: Thu Jul 1 21:44:28 CDT 1999 -->
<!-- hhmts start -->
Last modified: Fri Jul 2 11:24:48 CDT 1999
<!-- hhmts end -->
</body>
</html>
|
