I am working on a CXF based web service in Jboss Fuse environment. Is it okay if I use the default org.osgi.service.http.port=8181 defined in the pax.web.cfg or should I always specify an explicit port in my cxf endpoints?
For the first option, my cxf endpoint would look like:
<cxf:cxfEndpoint id="myWebService.endpoint"
address="/services/WebService/1.0"
...
For the second option, it would look like:
<cxf:cxfEndpoint id="myWebService.endpoint"
address="http://0.0.0.0:{port}/services/WebService/1.0"
...
Is it allowed to use port 8181 for all my CXF web services?
It is totally possible to use default service port for web service address. Looks like only when want to apply SSL, we should do that with httpj:engine settings for server side SSL.
Link to Fuse HTTPS settings
Related
I am adding the secure port with the non-secure port already opened. I wants to separate the traffic between two and forward to different Spring boot 2 Controllers. Wonder how I can achieve that?
In most solutions I have seen so far, https / SSL is terminated infront of Tomcat or Spring Boot application, so that Tomcat / Spring Controller receives only http requests on port 8080 (for example).
Termination of SSL in front of Tomcat / Spring Boot could be done with a Reverse Proxy or Web Server, like Apache2 or nginx.
Then the communication flow looks like this:
User ==HTTP-80==> Apache2 ==HTTP-8080==> Tomcat/Spring Boot
User ==HTTPS-443==> Apache2 ==HTTP-8080==> Tomcat/Spring Boot
("HTTP-80" means HTTP protocol on port 80. "== ==>" is arrow showing communication flow.)
I already have a tomcat server running in a VM with port as 443 and redirect port as 8443. Can I configure the redirect port for spring boot application also as 8443 and run in same VM? Would I face error like port already in use? If yes, are there any specific port to be used for this purpose? I would not like to try since this is a production environment VM.
By default the https port used is 9393 in springboot.So , in your application if you need it to be 8443 , you need to configure it in the application.properties or application.yml like
application.yml
server:
port: 8443
or
application.properties
server.port=8443
Yes, you will have issues if some other application is using the same https port on the same VM, you will have to find a port that is not used by any other application and assign it for your springboot application. Check in your vm if any other application is already mapped to this port, if not you can use this port without any issues.
Please note that :
If HTTPS is enabled, it will completely replace HTTP as the protocol
over which the REST endpoints and the Data Flow Dashboard interact.
Plain HTTP requests will fail - therefore, make sure that you
configure your Shell accordingly.
Spring doc.
For security purposes, I'd like to have javamelody running on my app management port, not the main traffic port. This is a 1.5.13 spring boot app running jetty with some actuator endpoints enabled. I have set the actuator property management.port to something other than the server.port.
I can run javamdelody in a standalone jetty server on a third port using an approach like this github project. But three ports is a little wieldy in a micro-service context.
This stackoverflow post has answers for Tomcat, but not for Jetty. Any suggestions to run javamelody on the management port with Jetty?
What kind of configuration should be used to make the apache ignite used in a restful webservice to identify itself the host and port wherever it is deployed.
I could see directly giving host and port name in the example-ignite.xml file.
Ignite doesn't know whether it's in web service or not. You need to configure your own web service, for example, using jetty + jersey and call Ignite API from it when it's needed:
Ignite Spring Boot integration example, Spring Boot REST service example
Ignite also comes with it's own REST API:
https://apacheignite.readme.io/docs/rest-api
Based on the documentation, the #EnableSidecar annotation acts as a proxy for non-JVM applications that wish to register in Eureka. Accordingly, the configuration to be set is:
sidecar:
port: 81 <-- Port of the non-JVM application
health-uri: http://10.135.16.50:${sidecar.port}/api/health.php <-- URI of the non-JVM application's health endpoint
home-page-uri: http://10.135.16.50:${sidecar.port}/
Once the "sidecar" is up and running, we should be able to invoke one of the non-JVM endpoints through the service registry just by using the name that the "sidecar" application used to register in Eureka. So for example, if our "sidecar" application was registered in Eureka as "php-sidecar" to proxy a PHP application with an endpoint such as:
http://10.135.16.50:81/api/microservice.php
Then we should be able to invoke the following endpoint to get to the non-JVM application (assuming "sidecar" is in "localhost" and port 8080):
http://localhost:8080/php-sidecar/api/microservice.php
However, this is not working as expected. When we request the URI just above, a request to "localhost:81" is actually issued, because somehow the "sidecar" is picking up its host URI and not the home-page-uri defined as part of the sidecar's properties.
If we run the non-JVM application locally using localhost, then everything works as expected, but this is definitively not the realistic use case.
Therefore, what am I missing in my configuration to tell Spring Cloud (Zuul in this particular case) to use the non-JVM home-page-uri and not my local host URI?
Thanks for your support.
After some research, it turns out that the sidecar must always be deployed in the same host as the non-JVM application.
After some research, it turns out that the sidecar must always be deployed in the same host as the non-JVM application.
No, You can add sidecar.ip-address = 10.135.16.50 to your yaml config file
I'm using spring-boot-starter-parent version is 1.5.6.RELEASE
working fine