File size: 25,459 Bytes
e8f1805 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 | # Aggregate-eip.md
The
[Aggregator](http://www.enterpriseintegrationpatterns.com/Aggregator.html)
from the [EIP patterns](#enterprise-integration-patterns.adoc) allows
you to combine a number of messages into a single message.
How do we combine the results of individual, but related, messages so
that they can be processed as a whole?
<figure>
<img src="eip/Aggregator.gif" alt="image" />
</figure>
Use a stateful filter, an Aggregator, to collect and store individual
messages until a complete set of related messages has been received.
Then, the Aggregator publishes a single message distilled from the
individual messages.
The aggregator is one of the most complex EIP and has many features and
configurations.
The logic for combing messages together is *correlated* in buckets based
on a *correlation key*. Messages with the same correlation key are
aggregated together, using an `AggregationStrategy`.
# Aggregate options
# Exchange properties
# Worker pools
The aggregate EIP will always use a worker pool used to process all the
outgoing messages from the aggregator. The worker pool is determined
accordingly:
- If a custom `ExecutorService` has been configured, then this is used
as worker pool.
- If `parallelProcessing=true` then a *default* worker pool (is 10
worker threads by default) is created. However, the thread pool size
and other configurations can be configured using *thread pool
profiles*.
- Otherwise, a single threaded worker pool is created.
- To achieve synchronous aggregation, use an instance of
`SynchronousExecutorService` for the `executorService` option. The
aggregated output will execute in the same thread that called the
aggregator.
# Aggregating
The `AggregationStrategy` is used for aggregating the old, and the new
exchanges together into a single exchange; that becomes the next old,
when the next message is aggregated, and so forth.
Possible implementations include performing some kind of combining or
delta processing. For instance, adding line items together into an
invoice or just using the newest exchange and removing old exchanges
such as for state tracking or market data prices, where old values are
of little use.
Notice the aggregation strategy is a mandatory option and must be
provided to the aggregator.
In the aggregate method, do not create a new exchange instance to
return, instead return either the old or new exchange from the input
parameters; favor returning the old exchange whenever possible.
Here are a few example `AggregationStrategy` implementations that should
help you create your own custom strategy.
//simply combines Exchange String body values using '+' as a delimiter
class StringAggregationStrategy implements AggregationStrategy {
public Exchange aggregate(Exchange oldExchange, Exchange newExchange) {
if (oldExchange == null) {
return newExchange;
}
String oldBody = oldExchange.getIn().getBody(String.class);
String newBody = newExchange.getIn().getBody(String.class);
oldExchange.getIn().setBody(oldBody + "+" + newBody);
return oldExchange;
}
}
//simply combines Exchange body values into an ArrayList<Object>
class ArrayListAggregationStrategy implements AggregationStrategy {
public Exchange aggregate(Exchange oldExchange, Exchange newExchange) {
Object newBody = newExchange.getIn().getBody();
ArrayList<Object> list = null;
if (oldExchange == null) {
list = new ArrayList<Object>();
list.add(newBody);
newExchange.getIn().setBody(list);
return newExchange;
} else {
list = oldExchange.getIn().getBody(ArrayList.class);
list.add(newBody);
return oldExchange;
}
}
}
The `org.apache.camel.builder.AggregationStrategies` is a builder that
can be used for creating commonly used aggregation strategies without
having to create a class.
## Aggregate by grouping exchanges
In the route below we group all the exchanges together using
`GroupedExchangeAggregationStrategy`:
from("direct:start")
// aggregates all using the same expression and group the
// exchanges, so we get one single exchange containing all
// the others
.aggregate(new GroupedExchangeAggregationStrategy()).constant(true)
// wait for 0.5 seconds to aggregate
.completionTimeout(500L).to("mock:result");
As a result we have one outgoing `Exchange` being routed to the
`"mock:result"` endpoint. The exchange is a holder containing all the
incoming Exchanges.
The output of the aggregator will then contain the exchanges grouped
together in a list as shown below:
List<Exchange> grouped = exchange.getMessage().getBody(List.class);
## Aggregating into a List
If you want to aggregate some value from the messages `<V>` into a
`List<V>` then you can use the
`org.apache.camel.processor.aggregate.AbstractListAggregationStrategy`
abstract class.
The completed Exchange sent out of the aggregator will contain the
`List<V>` in the message body.
For example, to aggregate a `List<Integer>` you can extend this class as
shown below, and implement the `getValue` method:
public class MyListOfNumbersStrategy extends AbstractListAggregationStrategy<Integer> {
@Override
public Integer getValue(Exchange exchange) {
// the message body contains a number, so return that as-is
return exchange.getIn().getBody(Integer.class);
}
}
The `org.apache.camel.builder.AggregationStrategies` is a builder that
can be used for creating commonly used aggregation strategies without
having to create a class.
The previous example can also be built using the builder as shown:
AggregationStrategy agg = AggregationStrategies.flexible(Integer.class)
.accumulateInCollection(ArrayList.class)
.pick(body());
## Aggregating on timeout
If your aggregation strategy implements
`TimeoutAwareAggregationStrategy`, then Camel will invoke the `timeout`
method when the timeout occurs. Notice that the values for index and
total parameters will be -1, and the timeout parameter will be provided
only if configured as a fixed value. You must **not** throw any
exceptions from the `timeout` method.
## Aggregate with persistent repository
The aggregator provides a pluggable repository which you can implement
your own `org.apache.camel.spi.AggregationRepository`.
If you need a persistent repository, then Camel provides numerous
implementations, such as from the
[Caffeine](#ROOT:caffeine-cache-component.adoc),
[CassandraQL](#ROOT:cql-component.adoc),
[EHCache](#ROOT:ehcache-component.adoc),
[Infinispan](#ROOT:infinispan-component.adoc),
[JCache](#ROOT:jcache-component.adoc), [LevelDB](#others:leveldb.adoc),
[Redis](#others:redis.adoc), or [SQL](#ROOT:sql-component.adoc)
components.
# Completion
When aggregation [Exchange](#manual::exchange.adoc)s at some point, you
need to indicate that the aggregated exchanges are complete, so they can
be sent out of the aggregator. Camel allows you to indicate completion
in various ways as follows:
- *completionTimeout*: Is an inactivity timeout in that is triggered
if no new exchanges have been aggregated for that particular
correlation key within the period.
- *completionInterval*: Once every X period all the current aggregated
exchanges are completed.
- *completionSize*: Is a number indicating that after X aggregated
exchanges its complete.
- *completionPredicate*: Runs a [Predicate](#manual::predicate.adoc)
when a new exchange is aggregated to determine if we are complete or
not. The configured aggregationStrategy can implement the Predicate
interface and will be used as the completionPredicate if no
completionPredicate is configured. The configured
aggregationStrategy can override the `preComplete` method and will
be used as the completionPredicate in pre-complete check mode. See
further below for more details.
- *completionFromBatchConsumer*: Special option for [Batch
Consumer](#manual::batch-consumer.adoc), which allows you to
complete when all the messages from the batch have been aggregated.
- *forceCompletionOnStop*: Indicates to complete all current
aggregated exchanges when the context is stopped
- *AggregateController*: which allows to use an external source
(`AggregateController` implementation) to complete groups or all
groups. This can be done using Java or JMX API.
All the different completions are per correlation key. You can combine
them in any way you like. It’s basically the first that triggers that
wins. So you can use a completion size together with a completion
timeout. Only completionTimeout and completionInterval cannot be used at
the same time.
Completion is mandatory and must be configured on the aggregation.
## Pre-completion mode
There can be use-cases where you want the incoming
[Exchange](#manual::exchange.adoc) to determine if the correlation group
should pre-complete, and then the incoming
[Exchange](#manual::exchange.adoc) is starting a new group from scratch.
The pre-completion mode must be enabled by the `AggregationStrategy` by
overriding the `canPreComplete` method to return a `true` value.
When pre-completion is enabled then the `preComplete` method is invoked:
/**
* Determines if the aggregation should complete the current group, and start a new group, or the aggregation
* should continue using the current group.
*
* @param oldExchange the oldest exchange (is <tt>null</tt> on first aggregation as we only have the new exchange)
* @param newExchange the newest exchange (can be <tt>null</tt> if there was no data possible to acquire)
* @return <tt>true</tt> to complete current group and start a new group, or <tt>false</tt> to keep using current
*/
boolean preComplete(Exchange oldExchange, Exchange newExchange);
If the `preComplete` method returns `true`, then the existing
correlation group is completed (without aggregating the incoming
exchange (`newExchange`). Then the `newExchange` is used to start the
correlation group from scratch, so the group would contain only that new
incoming exchange. This is known as pre-completion mode.
The `newExchange` contains the following exchange properties, which can
be used to determine whether to pre complete.
<table>
<colgroup>
<col style="width: 30%" />
<col style="width: 10%" />
<col style="width: 59%" />
</colgroup>
<thead>
<tr class="header">
<th style="text-align: left;">Property</th>
<th style="text-align: left;">Type</th>
<th style="text-align: left;">Description</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td
style="text-align: left;"><p><code>CamelAggregatedSize</code></p></td>
<td style="text-align: left;"><p><code>int</code></p></td>
<td style="text-align: left;"><p>The total number of messages
aggregated.</p></td>
</tr>
<tr class="even">
<td
style="text-align: left;"><p><code>CamelAggregatedCorrelationKey</code></p></td>
<td style="text-align: left;"><p><code>String</code></p></td>
<td style="text-align: left;"><p>The correlation identifier as a
<code>String</code>.</p></td>
</tr>
</tbody>
</table>
When the aggregation is in *pre-completion* mode, then only the
following completions are in use:
- *completionTimeout* or *completionInterval* can also be used as
fallback completions
- any other completions are not used (such as by size, from batch
consumer, etc.)
- *eagerCheckCompletion* is implied as `true`, but the option has no
effect
## CompletionAwareAggregationStrategy
If your aggregation strategy implements
`CompletionAwareAggregationStrategy`, then Camel will invoke the
`onComplete` method when the aggregated `Exchange` is completed. This
allows you to do any last minute custom logic such as to clean up some
resources, or additional work on the exchange as it’s now completed. You
must **not** throw any exceptions from the `onCompletion` method.
## Completing the current group decided from the AggregationStrategy
The `AggregationStrategy` supports checking for the exchange property
(`Exchange.AGGREGATION_COMPLETE_CURRENT_GROUP`) on the returned
`Exchange` that contains a boolean to indicate if the current group
should be completed. This allows to overrule any existing completion
predicates / sizes / timeouts etc., and complete the group.
For example, the following logic will complete the group if the message
body size is larger than 5. This is done by setting the exchange
property `Exchange.AGGREGATION_COMPLETE_CURRENT_GROUP` to `true`.
public final class MyCompletionStrategy implements AggregationStrategy {
@Override
public Exchange aggregate(Exchange oldExchange, Exchange newExchange) {
if (oldExchange == null) {
return newExchange;
}
String body = oldExchange.getIn().getBody(String.class) + "+"
+ newExchange.getIn().getBody(String.class);
oldExchange.getIn().setBody(body);
if (body.length() >= 5) {
oldExchange.setProperty(Exchange.AGGREGATION_COMPLETE_CURRENT_GROUP, true);
}
return oldExchange;
}
}
## Completing all previous group decided from the AggregationStrategy
The `AggregationStrategy` checks an exchange property, from the returned
exchange, indicating if all previous groups should be completed.
This allows to overrule any existing completion predicates / sizes /
timeouts etc., and complete all the existing previous group.
The following logic will complete all the previous groups, and start a
new aggregation group.
This is done by setting the property
`Exchange.AGGREGATION_COMPLETE_ALL_GROUPS` to `true` on the returned
exchange.
public final class MyCompletionStrategy implements AggregationStrategy {
@Override
public Exchange aggregate(Exchange oldExchange, Exchange newExchange) {
if (oldExchange == null) {
// we start a new correlation group, so complete all previous groups
newExchange.setProperty(Exchange.AGGREGATION_COMPLETE_ALL_GROUPS, true);
return newExchange;
}
String body1 = oldExchange.getIn().getBody(String.class);
String body2 = newExchange.getIn().getBody(String.class);
oldExchange.getIn().setBody(body1 + body2);
return oldExchange;
}
}
## Manually force the completion of all aggregated Exchanges immediately
You can manually trigger completion of all current aggregated exchanges
by sending an exchange containing the exchange property
`Exchange.AGGREGATION_COMPLETE_ALL_GROUPS` set to `true`. The message is
considered a signal message only, the message headers/contents will not
be processed otherwise.
You can alternatively set the exchange property
`Exchange.AGGREGATION_COMPLETE_ALL_GROUPS_INCLUSIVE` to `true` to
trigger completion of all groups after processing the current message.
## Using a controller to force the aggregator to complete
The `org.apache.camel.processor.aggregate.AggregateController` allows
you to control the aggregate at runtime using Java or JMX API. This can
be used to force completing groups of exchanges, or query its current
runtime statistics.
The aggregator provides a default implementation if no custom one has
been configured, which can be accessed using `getAggregateController()`
method. Though it may be easier to configure a controller in the route
using `aggregateController` as shown below:
private AggregateController controller = new DefaultAggregateController();
from("direct:start")
.aggregate(header("id"), new MyAggregationStrategy())
.completionSize(10).id("myAggregator")
.aggregateController(controller)
.to("mock:aggregated");
Then there is API on `AggregateController` to force completion. For
example, to complete a group with key foo:
int groups = controller.forceCompletionOfGroup("foo");
The returned value is the number of groups completed. A value of 1 is
returned if the foo group existed, otherwise 0 is returned.
There is also a method to complete all groups:
int groups = controller.forceCompletionOfAllGroups();
The controller can also be used in XML DSL using the
`aggregateController` to refer to a bean with the controller
implementation, which is looked up in the registry.
When using Spring XML, you can create the bean with `<bean>` as shown:
<bean id="myController" class="org.apache.camel.processor.aggregate.DefaultAggregateController"/>
<camelContext xmlns="http://camel.apache.org/schema/spring">
<route>
<from uri="direct:start"/>
<aggregate aggregationStrategy="myAppender" completionSize="10"
aggregateController="myController">
<correlationExpression>
<header>id</header>
</correlationExpression>
<to uri="mock:result"/>
</aggregate>
</route>
</camelContext>
There is also JMX API on the aggregator which is available under the
processors node in the Camel JMX tree.
# Aggregating with Beans
To use the `AggregationStrategy` you had to implement the
`org.apache.camel.AggregationStrategy` interface, which means your logic
would be tied to the Camel API. You can use a bean for the logic and let
Camel adapt to your bean. To use a bean, then a convention must be
followed:
- there must be a public method to use
- the method must not be void
- the method can be static or non-static
- the method must have two or more parameters
- the parameters are paired, so the first half is applied to the
`oldExchange`, and the reminder half is for the `newExchange`.
Therefore, there must be an equal number of parameters, e.g., 2, 4,
6, etc.
The paired methods are expected to be ordered as follows:
- the first parameter is the message body
- optional, the second parameter is a `Map` of the headers
- optional, the third parameter is a `Map` of the exchange properties
This convention is best explained with some examples.
In the method below, we have only two parameters, so the first parameter
is the body of the `oldExchange`, and the second is paired to the body
of the `newExchange`:
public String append(String existing, String next) {
return existing + next;
}
In the method below, we have only four parameters, so the first
parameter is the body of the `oldExchange`, and the second is the `Map`
of the `oldExchange` headers, and the third is paired to the body of the
`newExchange`, and the fourth parameter is the `Map` of the
`newExchange` headers:
public String append(String existing, Map existingHeaders, String next, Map nextHeaders) {
return existing + next;
}
And finally, if we have six parameters, that includes the exchange
properties:
public String append(String existing, Map existingHeaders, Map existingProperties,
String next, Map nextHeaders, Map nextProperties) {
return existing + next;
}
To use this with the aggregate EIP, we can use a bean with the aggregate
logic as follows:
public class MyBodyAppender {
public String append(String existing, String next) {
return next + existing;
}
}
And then in the Camel route we create an instance of our bean, and then
refer to the bean in the route using `bean` method from
`org.apache.camel.builder.AggregationStrategies` as shown:
private MyBodyAppender appender = new MyBodyAppender();
public void configure() throws Exception {
from("direct:start")
.aggregate(constant(true), AggregationStrategies.bean(appender, "append"))
.completionSize(3)
.to("mock:result");
}
We can also provide the bean class type directly:
public void configure() throws Exception {
from("direct:start")
.aggregate(constant(true), AggregationStrategies.bean(MyBodyAppender.class, "append"))
.completionSize(3)
.to("mock:result");
}
And if the bean has only one method, we do not need to specify the name
of the method:
public void configure() throws Exception {
from("direct:start")
.aggregate(constant(true), AggregationStrategies.bean(MyBodyAppender.class))
.completionSize(3)
.to("mock:result");
}
And the `append` method could be static:
public class MyBodyAppender {
public static String append(String existing, String next) {
return next + existing;
}
}
If you are using XML DSL, then we need to declare a `<bean>` with the
bean:
<bean id="myAppender" class="com.foo.MyBodyAppender"/>
And in the Camel route we use `aggregationStrategy` to refer to the bean
by its id, and the `strategyMethodName` can be used to define the method
name to call:
<camelContext xmlns="http://camel.apache.org/schema/spring">
<route>
<from uri="direct:start"/>
<aggregate aggregationStrategy="myAppender" aggregationStrategyMethodName="append" completionSize="3">
<correlationExpression>
<constant>true</constant>
</correlationExpression>
<to uri="mock:result"/>
</aggregate>
</route>
</camelContext>
When using XML DSL, you can also specify the bean class directly in
`aggregationStrategy` using the `#class:` syntax as shown:
<route>
<from uri="direct:start"/>
<aggregate aggregationStrategy="#class:com.foo.MyBodyAppender" aggregationStrategyMethodName="append" completionSize="3">
<correlationExpression>
<constant>true</constant>
</correlationExpression>
<to uri="mock:result"/>
</aggregate>
</route>
You can use this in XML DSL when you are not using the classic Spring
XML files ( where you use XML only for Camel routes).
## Aggregating when no data
When using bean as `AggregationStrategy`, then the method is **only**
invoked when there is data to be aggregated, meaning that the message
body is not `null`. In cases where you want to have the method invoked,
even when there are no data (message body is `null`), then set the
`strategyMethodAllowNull` to `true`.
When using beans, this can be configured a bit easier using the
`beanAllowNull` method from `AggregationStrategies` as shown:
public void configure() throws Exception {
from("direct:start")
.pollEnrich("seda:foo", 1000, AggregationStrategies.beanAllowNull(appender, "append"))
.to("mock:result");
}
Then the `append` method in the bean would need to deal with the
situation that `newExchange` can be `null`:
public class MyBodyAppender {
public String append(String existing, String next) {
if (next == null) {
return "NewWasNull" + existing;
} else {
return existing + next;
}
}
}
In the example above we use the [Content
Enricher](#content-enricher.adoc) EIP using `pollEnrich`. The
`newExchange` will be `null` in the situation we could not get any data
from the "seda:foo" endpoint, and a timeout was hit after 1 second.
So if we need to do special merge logic, we would need to set
`setAllowNullNewExchange=true`. If we didn’t do this, then on timeout
the append method would normally not be invoked, meaning the [Content
Enricher](#content-enricher.adoc) did not merge/change the message.
In XML DSL you would configure the `strategyMethodAllowNull` option and
set it to `true` as shown below:
<camelContext xmlns="http://camel.apache.org/schema/spring">
<route>
<from uri="direct:start"/>
<aggregate aggregationStrategy="myAppender"
aggregationStrategyMethodName="append"
aggregationStrategyMethodAllowNull="true"
completionSize="3">
<correlationExpression>
<constant>true</constant>
</correlationExpression>
<to uri="mock:result"/>
</aggregate>
</route>
</camelContext>
## Aggregating with different body types
When, for example, using `strategyMethodAllowNull` as `true`, then the
parameter type of the message bodies does not have to be the same. For
example suppose we want to aggregate from a `com.foo.User` type to a
`List<String>` that contains the name of the user. We could code a bean
as follows:
public final class MyUserAppender {
public List addUsers(List names, User user) {
if (names == null) {
names = new ArrayList();
}
names.add(user.getName());
return names;
}
}
Notice that the return type is a `List` which we want to contain the
name of the users. The first parameter is the `List` of names, and the
second parameter is the incoming `com.foo.User` type.
|