We're using Worklight 6.1.0.0 / WebSphere 8.0.0.2 (ND/aix).
This seemed pretty close to my question too, but for version 6.0.
I've successfully done uninstall/install to our worklight console war package. However, there is some extra work on re-deploying adapters and such. I was looking for a way to just update the console. Among the ant tasks there is a target 'minimal-update', which sounds like what I'm looking for (is it?). However when all other pieces fell into place, I have an error for mapping the datasources:
ADMA0007E: A validation error occurred in task Mapping resource references to resources. The Java Naming and Directory Interface (JNDI) name is not specified for resource reference jdbc/WorklightDS in module Worklight with EJB name .
Contents of the 'minimal-update' task is pretty much the same as for 'install'.
I tried that as update from websphere admin console (but i should use the ant task - right?), that gave me a wizard screen to map jdbc/WorklightDS from package to jdbc/WorklightDS on server. This left me wondering how could I tell this using the ant task.
The ant target minimal-update of the sample configuration files documented at http://pic.dhe.ibm.com/infocenter/wrklight/v6r1m0/topic/com.ibm.worklight.deploy.doc/devref/c_ant_tasks_sample_config_files.html is meant to update a WAR file already deployed (and not uninstalled). In particular, on WAS, it assumes that the JNDI datasources are in place.
If you have uninstalled the WAR file, you should use the target install instead, provided that your databases were created for Worklight 6.1.
If they were created for a previous version of worklight you must upgrade their schema as well running the target 'databases' (and if it's a production installation, you might want to read all the steps in detail at http://pic.dhe.ibm.com/infocenter/wrklight/v6r1m0/topic/com.ibm.worklight.upgrade.doc/devenv/c_upgrade_to_srvr610_in_production_env.html )
Related
I'm working on an IBM MobileFirst Platform 7.1 project where there are many name value pairs in the worklight.properties file.
Say for example,
HOST_NAME = www.google.com
I access these variables form the MobileFirst adapter.
var hostName = WL.Server.configuration["HOST_NAME"]
There will be cases where I'll have to change the HOST_NAME in the worklight.properties to a different value. But sometimes, even after the value is changed in the worklight.properties file, they are not getting updated in the adapter.
Every time when I change the values I do the following,
Clean the project
Restart the server
Deploy the adapter again
Nothing seemed to change the value.
I suspect, the MobileFirst server caches the values and reuses the old values without updating the new ones. I say this because, I tried deleting the values in the worklight.properties file, and even tried deleting the file itself. Still the adapter was using the old value.
I tried deleting the MobileFirstServerConfig Folder in my workspace. Even that didn't work.
It would be great if someone could help me clear the cache or find a work around to this issue. Thanks in advance.
I kind of followed the above method but initially it didn't work.
But later I tried to follow these steps in this exact order and it seemed to work.
Stop Worklight Server.
Remove your project from Mobile First Development Server.
Go to the "bin" directory and remove your project war.
Clean and Build the entire workspace.
Start the server
Deploy the adapters
Run your app.
The worklight.properties file is embedded into the project runtime WAR file. So, if you change anything in worklight.properties, you must rebuild the WAR file and redeploy it. Based on your description, it sounds like you're only building the adapter (and not the app). At minimum, I would suggest to remove the runtime WAR file from the project "bin" directory (just to ensure that it gets rebuilt) and build all app environments. That should build / redeploy the WAR file and restart the server, and then what you're doing should work.
Here are the steps to follow:
1. If worklight server is up and running, stop it or restart your system.
2. Clear bin folder.
3. Do Clean Project and Build for enviournment.
4. Open Mobile first Console(Start server)
5. Deploy All adapters
6. Run your app.
If wlq or wl error is comming we can give an alert message 'test' in our main file where we have all the javascript files.
We have a worklight i/OS hybrid application built and deployed to the worklight console using RTC jazz team build and the ant-builder ant tasks (6.1.0.1). The ipa packaging is done on a separate machine, though using the same level of Worklight Studio (6.1.0.1).
I've compared the wlapp files that are generated, and they are basically identical except for some whitespace characters (tabs, line feeds), probably due to the different environments (Windows vs AIX), and the following:
index.html
"WORKLIGHT_NATIVE_VERSION": "3921556017",
"WORKLIGHT_PLATFORM_VERSION": "6.1.0.01.20140311-2356",
"WORKLIGHT_NATIVE_VERSION": "1475155033",
"WORKLIGHT_PLATFORM_VERSION": "6.1.0.01.20140311-2356",
deployment.data
native=3921556017
native=1475155033
However, the direct update never happens when the wlapp is updated on the server.
on the WL.Client.connect call, the following json comes back as part of the response
gadgetProps":{"directUpdate":{},"ENVIRONMENT":"iphone"}
What are we missing here? What can be preventing the direct update from triggering?
Any help would be greatly appreciated.
The 6.1.0.1 iFix version 6.1.0.01.20140518-1532 from IBM Fix Central seems to have resolved this problem for us.
And what if you are not using ant, but rather built the project in Worklight Studio straight to Xcode from then install the .ipa and test Direct Update. Does it work?
From your question it is not clear whether or not you've confirmed one or the other.
Regardless, since the v6.1.0.1 build you are using, several Direct Update-related fixes were introduce; one with close proximity to the error you mention (directUpdate:{ }), so I suggest to to open a PMR in order to receive the latest available iFix (not yet available at IBM Fix Central).
I have a Worklight 6.0 project that uses the new Dojo 1.9 libs, I created an external dojo project, like the documentation suggested, then, in the main project properties, under "Dojo toolkit", it references this dojo19 project.
The project works on the local server, then I did "Run As" | "Build for Remote Server...", and entered the correct domain:port and context path, clicked Build, the *.wlapp files were updated. (I've also updated the settings for publicWorkLightHostname / publicWorkLightPort / publicWorkLightProtocol in the "Environment Entries for Web Modules" in the installed war to match the remote server names/port/protocol.)
But, after deploying both war and -all.wlapp file, accessing the app I get JS errors when it tries to refer to the dojo19 library:
The page at
https://<myIP>:9443/<myproject>/apps/services/www/ /mobilewebapp/default/IODMobile.html
ran insecure content from http://localhost:64441/dojo19/<myproject>/IODMobile/mobilewebapp/dojo/nls/core-web-layer_en-us.js.
The dojo19 is the project name in my Worklight developer workspace that I referred to above.
Why is it trying localhost? Seems there's a missing step here in deploying the dojo library project into Worklight.
Where are you trying to preview the application when you get the error message?
See the changes in Dojo in Worklight 6.0
If launching the application in emulator/simulator/device, see Billy Rowe's answer in this question
Partial copy/paste:
Step 1: Verify your application works in the Mobile Browser Simulator
with Provide Library Resources checked. If the Console log is showing
resources being served from the server, then these have to be copied
to your application before deploying to AVD or a device
Step 2: After you think you have all Dojo/resources within your
project, uncheck Provide Library Resources and test it again in MBS.
If it fails in MBS, then something is missing in your application that
is in the library/server. You can check Provide Library Resources and
retest to see if it shows you what that is. Not all resources are
shown, e.g. if there's a missing CSS file.
Also I would suggest to do all of this in the Development environment (that is, in Eclipse) before starting to deploy the .war file and .wlapp file etc... (which, BTW, I hope you're doing based on the new instructions for Worklight 6.0)
In the information center, it will show you how to uncheck the Provide Library Resources in the Console Log.
I think what you're running into is:
1) Something is being served from the Dojo Library/Server
2) A bug in 6.0 that used "localhost" instead of the IP of the host (your machine running eclipse). This is fixed in the 6.0 iFix. With this fix, you can run your app external to Studio and still use the Dojo Library/Server. Without this fix, you must have everything you need within your app.
Can you install the iFix and let us know if that fixed the problem?
I want to enable user customization at run time with Oracle ADF. With JDeveloper, if I deploy my application directly to the server, a window permits me to choose a repository previously registered on my Enterprise Manager, as depicted in following image:
In this way my application works great. Now I would to deploy to EAR file (in order to save a copy of my release) but I don't know how to target MDS repository. On the web I've found that maybe the adf-config.xml file has to be modify, but in which way?
have a look at this blog entry, which explains how to do it. Apparently this information needs to be provided upon EAR file deployment
http://andrejusb.blogspot.de/2011/05/target-mds-repository-for-adf.html
I have a 2 war projects deployed as a library on a weblogic server (version 1 and version 2) and an EAR project deployed on the same server that uses the library.
When I deploy the library version 2 stays on prepared state and the version 1 remains in active state. Does somebody know how can I do to make active the 2nd version of my war?
thanks!
Under deployments, did you try selecting the library of interest and explicitly clicking on Activate? If you do not see the library there, you may have to modify the view of your console to include libraries in your deployment page.
If there are conflicts with the other library, you'll get the full stack trace within the stand log files. You also have the option of stopping the currently 'active' library before enabling the newer one to avoid conflicts.