I was building a new Spring WebFlux application with Spring Boot. After downloading the project template from start.spring.io, I added some third-party dependencies and tried to start the application. Then I met the error org.springframework.context.ApplicationContextException: Unable to start ServletWebServerApplicationContext due to missing ServletWebServerFactory bean. Actually this is a common issue when I searched the solution. All the solutions are related to missing dependencies of Spring Boot starter spring-boot-starter-web.
When building large-scale NodeJS and front-end projects, it’s necessary to use private NPM registry to share private packages across different projects and team members. This post shows the essential parts of how to build the infrastructure and set up the workflow to use them.
It’s a bit tricky to enable JMX for Cassandra when running in Docker. I create a Cassandra Docker image alexcheng/cassandra of version 3.11.2 with JMX authentication enabled, see the Dockerfile. The username and password to access JMX are both cassandra. The environment variable JMX_HOST has the default value of localhost, and it should work most of the time. If localhost doesn’t work , you can use the host’s IP address. After starting the container, you can access JMX using VisualVM or JConsole with address service:jmx:rmi://localhost/jndi/rmi://localhost:7199/jmxrmi. Change localhost to the same value as JMX_HOST if localhost is not used. Since SSL is not enabled, you need to use insecure connections.
On Mac, the local IP address can be obtained using ipconfig getifaddr en0, so we can use following command to start the container when using IP address is required.
docker run --rm -p 7199:7199 -e "JMX_HOST=`ipconfig getifaddr en0`" alexcheng/cassandra
Atlassian got its new fancy tool Stride to replace HipChat and it’s gradually migrate existing HipChat teams to Stride. For those Hubots connecting to HipChat, it’s time to migrate. This post shows the complete migration process for my own ChatOps bot.
If your Java applications are deployed to AWS Elastic Beanstalk, you may want to configure the logging to make error diagnostics much easier. We can use Beanstalk console to request and download log files, but you need to configure the app first to make your log files available for download.
It’s common to use private Maven repositories to host internal artifacts. These repositories usually require authentication when accessing. The username and password for these repositories are stored in the ~/.m2/settings.xml. This post shows how can we configure Gradle to publish to private repositories and integrate with Bitbucket Pipelines or other CI services.
Akka Streams and Alpakka provide a good an alternative to Apache Camel when integrating different services and data. Alpakka has a rich set of connectors, including file, JMS, Cassandra and more. We can still use Camel endpoints if Alpakka doesn’t provide the connectors.
This post shows a complete example of integrating Camel endpoints with Akka Streams. This example reads file content in one directory, tranforms the content, then writes to another directory. It uses the Camel file endpoint.
Upgrading React Native should be quite easy. But if you have an app that uses a very old version of React Native, things may be different. This post shows how to upgrade React Native from 0.42 to 0.51.