I am considering using capedwarf to migrate some of my customers to their private servers, instead of on the app engine cloud. I already tried some of the scenarios that are mandatory in my applications, and I have been greatly surprised by the quality of the tool out of the box.
However, here are the issues I am facing, and I can't get to work:
1 - Session management on Chrome:
I have no issues connecting with firefox, but using Chrome, it seems that the HTTP session is lost somewhere. It seems something about the JSESSIONID is broken. I have tried using the previous version of capedwarf, and apparently, the bug has been introduced lately since the previous version are working as expected... Is there anyone else meeting the issue?
2 - Query cursors:
My application makes an intensive usage of the query cursors over the objectify library. Apparently, for a reason I didn't track, the cursor are totally ignored, and the results of all of my queries give the same results. I tried in previous version, and I found the same problem. Is it something not implemented?
3 - Remote API:
I didn't have any chances to make the remote API to work. I have a home made java application that uses the remote API for backup / recovery. I tried it without modifications, and could not connect to the application on wildfly using it. I have seen somewhere that it should be supported, but didn't find any documentations on the required steps to perform. Is there anything special to know?
Then, I am wondering about the configuration I need to create for those customers. Considering the fact that my application is dedicated to companies and that each instance of the application will be used by 50~ 10K persons.
Is the wildfly filesystem storage trustable enough for such a scenario? Should I consider moving to MongoDB or Cassandra?
I checked on how to use mongoDB for example, but didn't find any good tutorial or documentations on the topic (I am totally new to JBoss). Is there any places where I can find explanations on how to configure it?