String objToString = Integer.toHexString(System.identityHashCode(objectReference));
In order from strongest to weakest:
- Strong reference, e.g.
StringBuilder sb = new StringBuilder(); - Soft reference, e.g.
SoftReference<Widget> softRef = new SoftReference<>();- good for cache, tend to stick about while there's plenty of memory available - Weak reference, e.g.
WeakReference<Widget> weakRef = new WeakReference<>();- referant is cleared by GC when no more strong references detected - Phantom reference, e.g.
PhantomReference<Widget> phantomRef = new PhantomReference<>();- theget()method always returns null. Can be used to determine when an object is GC'd.. in reality rarely used.
Soft and weak references are quite similar. Phantom references are quite different!
DO NOT call close on the unwrapped connection - only on the delegate: http://stackoverflow.com/questions/28002176/tomcat-connection-pool-few-methods-not-releasing-connection
JdbcTemplateseems to be suppressing calls to close the connection when the connection has been unwrapped or the simple native JDBC extractor has been used for the same purpose. Wrapping the delegate connection in a try-with forces closure of the wrapping connection and this seems to work.
To declare and populate a map in a single statement, the following should work, but produces an incompatible types compiler error in the GEC (OK in Eclipse / JDK 8):
private static final Map<Level, Severity> severityMap = unmodifiableMap(Stream
.of(new SimpleImmutableEntry<>(Level.ERROR, Severity.MAJOR),
new SimpleImmutableEntry<>(Level.WARN, Severity.MINOR))
.collect(Collectors.toMap(Map.Entry::getKey,
Map.Entry::getValue)));
Instead we have to explicitly declare the type of the Strem and the collection operation:
private static final Map<Level, Severity> severityMap = unmodifiableMap(Stream
.<Entry<Level, Severity>> of(new SimpleImmutableEntry<>(Level.ERROR, Severity.MAJOR),
new SimpleImmutableEntry<>(Level.WARN, Severity.MINOR))
.collect(Collectors.<Entry<Level, Severity>, Level, Severity> toMap(Map.Entry::getKey,
Map.Entry::getValue)));
Bah! Rubbish.
In addition to the new stream-related methods, there are a number of really useful new default methods on the Map interface, for example:
computeIfAbsent(K key, Function<? super K,? extends V> mappingFunction);
computeIfPresent(K key, BiFunction<? super K,? super V,? extends V> remappingFunction);
putIfAbsent(K key, V value);etc - see Java 8 Map API
Some nice stuff: https://dzone.com/articles/java-8-type-annotations. In conjunction check out the Checker Framework
When using LongAccumulator in a multi-threaded environment always ensure that the binary accumulation function provided is not order-dependent. For example, Math::addExact would produce consistent results, where as Math::subtractExact would not (addition is order-independent, subtraction is not).
Great article working through and explaining different divide-and-conquer approaches to prime number generation, from single-thread brute through recursive fork-join to streams and spliterators: DZone article
For extensions to the Stream API providing for example some useful sequential stream functionality, check out the jOOL library: jOOL blog, jOOL API v0.9.7
InfoQ article e.g. StampedLock, LongAdder (always favour over AtomicLong, AtomicInteger etc due to high-contention performance improvements), Arrays.parallelSort(myArray), StringJoiner, Long.hashCode(long value)
- installed
- set to load on startup - to change go to Window - Preferences - General - Startup & Shutdown and untick all but the 'essential' ones.
Window - Preferences - General - Workspace - Local History
Some nice stuff esp. related to lambdas: https://www.eclipse.org/eclipse/news/4.5/jdt.php e.g. Quick assist (CTRL+1) for options to convert anonymous blocks to lambdas, lambdas to method references, make inferred lambda type parameters explicit, etc.
Regularly export preferences and code format (whenever a change has been made and adopted).
Add regularly used static imports to Window - Preferences - Java - Editor - Content Assist - Favourites. Good candidates are for example the Spring helper classes used in testing, e.g. MockRestResponseCreators.*, MockMvcResultMatchers.*, JUnit's Assert.* class, etc
Fix it:
git config --replace-all core.editor notepad
See https://multiplestates.wordpress.com/2015/02/05/rename-a-local-and-remote-branch-in-git/
-
Rename your local branch. If you are on the branch you want to rename:
git branch -m new-name
-
Delete the old-name remote branch and push the new-name local branch:
git push origin :old-name new-name
-
Reset the upstream branch for the new-name local branch. Switch to the branch and then:
git push origin -u new-name
Right click the origin server under Remotes -> origin and select 'Fetch'. Then checkout into a (new) local branch and switch to the new branch
<dependency>
<groupId>com.fasterxml.jackson.core</groupId>
<artifactId>jackson-core</artifactId>
<version>2.5.1</version>
</dependency>
<dependency>
<groupId>com.fasterxml.jackson.core</groupId>
<artifactId>jackson-databind</artifactId>
<version>2.5.1</version>
</dependency>
<dependency>
<groupId>com.fasterxml.jackson.core</groupId>
<artifactId>jackson-annotations</artifactId>
<version>2.5.0</version>
</dependency>Refer to: http://stackoverflow.com/questions/5816688/reseting-remote-to-a-certain-commit
git reset --hard <commit-hash>
git push -f origin master
This will discard previous commit so be careful when others may be working on the same remote branch!
Refer to: http://stackoverflow.com/questions/927358/how-do-you-undo-the-last-commit
git reset --soft HEAD~1
git push -f origin [name of branch]
That'll reset the active local branch.. then to push to the remote branch if you've really messed things up:
git merge -s recursive --strategy-option=theirs upstream/master
This will resolve all conflicts in favour of 'theirs'
After making a series of small or subsequently reverted commits, for example when iteratively testing some change affecting or affected by the CI builds, it is a good idea to squash all these meaningless commits into one single commit with the result being a sensible set of changes that reflect the feature being built. This can be done via an interactive rebase on the working branch:
git rebase -i <last_commit_to_preserve_as_is>
Then use the interactive editor to replace pick with squash or fixup for the second and subsequent commits you want
to squash into the first. Then optionally force-push the rewritten history to the remote tracking branch:
git push -f origin <branch_name>
sudo tcpdump -A -s 0 'tcp dst port 80'
sudo tcpdump -A -s 0 -i lo 'tcp dst port 8083'
(where tomcat is proxied on port 8083 via Apache). Crucial thing here is the -i lo to specify the LOCAL interface, not eth0, eth1 etc
There isn't a built in Hamcrest Matcher for this, so we implement one in a Groovy closure:
.andExpect(jsonPath('$.my.date.time', [
matches: { actual ->
try{
DateTimeFormatter.ofPattern("yyyy-MM-dd'T'HH:mm:ss.SSSZ").parse(actual.toString())
true
} catch(any){
any.printStackTrace()
false
} },
describeTo: { Description description -> 'should be a datetime' }
] as BaseMatcher))
The matcher returns true if the value can be parsed without exception, false if not.
Class 1 (do something in the background..):
@Component
public class BackgroundTask {
private int counter;
@Async
void slowCount() throws InterruptedException {
while (true) {
out.println("Counter is now: " + ++counter);
Thread.sleep(5000);
}
}
}Class 2: task scheduler
@Configuration
@EnableAsync
public class BackgroundTaskInitialiser {
@Autowired
private BackgroundTask backgroundTask;
@PostConstruct
public void initialise() throws InterruptedException {
backgroundTask.slowCount();
}
}Reference: http://docs.spring.io/spring/docs/current/spring-framework-reference/html/scheduling.html
In Run Configuration - vmargs:
-javaagent:"C:\applications\springloaded-1.2.1.RELEASE.jar" -noverify
When adding your own auto-config in the test harness this sort of thing can be done:
@SpringBootApplication
@ComponentScan("uk.gov.meto.commons.hal.api.testsupport.controller")
public class TestAppStarter {
/* required beans and stuff */
}That way, any other @Configuration classes will be located, but component scanning (e.g. for @ControllerAdvice annotated classes) will be limited to the specified base classes. This is handy for example when @ControllerAdvice classes don't have a default constructor but instead rely on beans to be created in the factory prior.
Or to run with maven spring-boot plugin, see: http://docs.spring.io/spring-boot/docs/current/reference/html/howto-hotswapping.html
Great comment on semantic diffs between two EIP frameworks: http://callistaenterprise.se/blogg/teknik/2015/10/12/apache-camel-vs-spring-integration/
Use an inbound http-gateway to setup a replying HTTP messaging endpoint akin to an MVC controller request mapping, and or an inbound channel when the request is one-way (response not required beyond 200 OK status returned):
@Bean
public IntegrationFlow locationsGateway() {
return IntegrationFlows
.from(inboundGateway("/accounts").requestMapping(r -> r.params("account_num"))
.payloadExpression("#requestParams.account_num"))
.enrichHeaders(headerSpec -> headerSpec.header(HttpHeaders.CONTENT_TYPE,
MediaType.APPLICATION_JSON_VALUE, Boolean.TRUE))
.gateway(obsLocationsOutboundFlow()).get();
}Message channels represent the 'pipes' in a pipes-and-filters architecture. Producers send messages to a channel and consumers recive messages from a channel. A message channel may follow either P2P (aka queue, at most one subscriber receives each msg) or Pub-Sub (aka topic; msgs broadcast to all subscribers) semantics.
In SI, *Pollable channels are capable of buffering messages in a queue, supporting throttling and overloading of a msg consumer.
mvn jar:jar install:install deploy:deploy
Run maven compile to resolve
Your workspace is likely corrupt - create a new one!
Quick fix - ignore - solved. Or: http://stackoverflow.com/questions/15938466
mvn archetype:create-from-project -DoutputDirectory ../pwms-springboot-service
:set fileformat=unix
Query xml with xpath from shell:
var1=$(xmllint --xpath "//*[local-name()='definitions']/*[local-name()='portType']/@name" service.wsdl)
Clear contents of: "C:\Users[user_name]\AppData\Local\Google\Chrome\User Data\Default\Cache"
Wrap the block in:
{code}Reading a CSV once an hour from an HTTP endpoint (via a POST request), unmarshalling it (lazily), splitting the contents by line then streaming the marhsalled JSON objects (each representing a line in the CSV) to the local logging system and a Kafka topic.
void configure() {
final CsvDataFormat csvFormat = new CsvDataFormat();
csvFormat.setUseMaps(TRUE); // unmarshal CSV to maps rather than Lists
csvFormat.setLazyLoad(TRUE); // read line-by-line (avoid OOM)
final String localKafkaTopicURI =
new StringJoiner("&").add("kafka:localhost:9092?topic=obs").add("zookeeperConnect=localhost:2181")
.add("serializerClass=kafka.serializer.StringEncoder").add("requestRequiredAcks=-1").toString();
from("timer://obstimer?fixedRate=true&period=1h").setHeader(Exchange.HTTP_METHOD, constant("POST"))
.setHeader(Exchange.CONTENT_TYPE, constant("application/x-www-form-urlencoded"))
.setBody(
simple("Type=Observation&PredictionSiteID=ALL&ObservationSiteID=ALL&Date=26%2F11%2F2015&PredictionTime=0500"))
.to("https://weather.data.gov.uk/query").routeId("obs").unmarshal(csvFormat).split(body()).streaming()
.marshal().json(JsonLibrary.Jackson).convertBodyTo(String.class)
.to("log:example.com.routers.landobs?level=TRACE&showExchangePattern=false").to(localKafkaTopicURI)
.end();
}In PAYLOAD data format, the raw XML from the transport layer containg the SOAP envelope has to be captured in a CXF interceptor before it is processed by CXF. The raw XML can then be stored on the CXF Message headers and retrieved later in the routing logic via the CxfConstants.CAMEL_CXF_MESSAGE exhange header object which contains the message.
See http://camel.465427.n5.nabble.com/Getting-entire-Soap-Message-with-header-and-body-in-Payload-mode-td5753162.html for example
sudo docker exec -it <containerID> bash
or, e.g. for Alpine linux:
sudo docker exec -it <containerID> /bin/sh
docker rm $(docker ps -a -q -f status=exited)
docker stats \`docker ps | awk '{print $NF}' | grep -v NAMES\`
In the background, with max container memory, host/guest OS port and volume mapping:
docker run --name my-service -d -m 256m -p 8080:8080 -v /data/my-service:/app/data tommyb/my-service
docker exec -it "id of running container" bash
$ aws s3api list-objects --bucket my-bucket --no-paginate --query "sort_by(Contents, &LastModified)[*].[Key,LastModified,Size]"
$ aws ec2 describe-tags --region eu-west-1 --filters "Name=resource-id,Values=vol-123456"
$ aws ec2 create-tags --region eu-west-1 --resources vol-123456 --tags "Key=Name,Value=BigVolume"
Different options described here: https://blog.cloudability.com/two-solutions-for-tagging-elusive-aws-ebs-volumes/
$ aws ec2 --region eu-west-1 describe-volumes --filter "Name=attachment.instance-id,Values=i-abc123" --query "Volumes[].VolumeId" --out text