Details

      Description

      Change the dependencies to Apache River.

        Activity

        Hide
        bryanthompson bryanthompson added a comment -

        Changed to Apache River 2.2. This looks like it will be a smooth
        transition. There also appears to be more jars than were present in
        jini 2.1.


        - AbstractFederation#startDeferredTasks(). Modified to trap an
        IllegalStateException and, if the federation was concurrently
        closed, log @ WARN and return rather than throwing out the
        exception.


        - Remove the jini 2.1 jars.


        - Added the river 2.2 jars. There are new jars in river which were
        not part of the jini 2.1 distribution. The new jars only appear in
        the lib directory (versus the lib-ext and lib-dl directories). The
        new jars are: asm-3.2.jar, asm-common-3.2.jar,
        checkconfigurationfile.jar, checkser.jar, classdep.jar,
        computedigest.jar, computehttpmdcodebase.jar, destroy.jar,
        envcheck.jar, extra.jar, group.jar, jarwrapper.jar,
        jsk-debug-policy.jar, outrigger-snaplogstore.jar,
        phoenix-group.jar, phoenix-init.jar, phoenix.jar,
        preferredlistgen.jar, serviceui.jar, sharedvm.jar.


        - I was able to run TestBigdataClient starting the individual
        services per the notes at
        https://sourceforge.net/apps/trac/bigdata/ticket/297.


        - No changes were required to build.xml or build.properties.


        - Renamed the license files for apache zookeeper and apache river to
        clarify their relationship to the jars. I have not renamed the
        directory structure within the bigdata-jini project as that would
        any number of problems with the deployment scripts.

        Committed revision r5254. (TERMS_REFACTOR_BRANCH)

        This change also needs to be applied to the 1.0.x maintenance branch. I want to watch the new jars in CI before applying the change to the 1.0.x maintenance branch as well.

        Show
        bryanthompson bryanthompson added a comment - Changed to Apache River 2.2. This looks like it will be a smooth transition. There also appears to be more jars than were present in jini 2.1. - AbstractFederation#startDeferredTasks(). Modified to trap an IllegalStateException and, if the federation was concurrently closed, log @ WARN and return rather than throwing out the exception. - Remove the jini 2.1 jars. - Added the river 2.2 jars. There are new jars in river which were not part of the jini 2.1 distribution. The new jars only appear in the lib directory (versus the lib-ext and lib-dl directories). The new jars are: asm-3.2.jar, asm-common-3.2.jar, checkconfigurationfile.jar, checkser.jar, classdep.jar, computedigest.jar, computehttpmdcodebase.jar, destroy.jar, envcheck.jar, extra.jar, group.jar, jarwrapper.jar, jsk-debug-policy.jar, outrigger-snaplogstore.jar, phoenix-group.jar, phoenix-init.jar, phoenix.jar, preferredlistgen.jar, serviceui.jar, sharedvm.jar. - I was able to run TestBigdataClient starting the individual services per the notes at https://sourceforge.net/apps/trac/bigdata/ticket/297 . - No changes were required to build.xml or build.properties. - Renamed the license files for apache zookeeper and apache river to clarify their relationship to the jars. I have not renamed the directory structure within the bigdata-jini project as that would any number of problems with the deployment scripts. Committed revision r5254. (TERMS_REFACTOR_BRANCH) This change also needs to be applied to the 1.0.x maintenance branch. I want to watch the new jars in CI before applying the change to the 1.0.x maintenance branch as well.
        Hide
        bryanthompson bryanthompson added a comment -

        I've run through the cluster install for 2 nodes with River on the dev-branch and the NSS and TCK test suites and all looks good.

        Changed to Apache River 2.2. This looks like it will be a smooth
        transition. There also appears to be more jars than were present in
        jini 2.1.


        - Remove the jini 2.1 jars.


        - Added the river 2.2 jars. There are new jars in river which were
        not part of the jini 2.1 distribution. The new jars only appear in
        the lib directory (versus the lib-ext and lib-dl directories). The
        new jars are: asm-3.2.jar, asm-common-3.2.jar,
        checkconfigurationfile.jar, checkser.jar, classdep.jar,
        computedigest.jar, computehttpmdcodebase.jar, destroy.jar,
        envcheck.jar, extra.jar, group.jar, jarwrapper.jar,
        jsk-debug-policy.jar, outrigger-snaplogstore.jar,
        phoenix-group.jar, phoenix-init.jar, phoenix.jar,
        preferredlistgen.jar, serviceui.jar, sharedvm.jar.


        - No changes were required to build.xml or build.properties.


        - Renamed the license files for apache zookeeper and apache river to
        clarify their relationship to the jars. I have not renamed the
        directory structure within the bigdata-jini project as that would
        any number of problems with the deployment scripts.


        - Updated the Depends.java file (in the TERMS branch also).

        Committed revision r5255.

        Show
        bryanthompson bryanthompson added a comment - I've run through the cluster install for 2 nodes with River on the dev-branch and the NSS and TCK test suites and all looks good. Changed to Apache River 2.2. This looks like it will be a smooth transition. There also appears to be more jars than were present in jini 2.1. - Remove the jini 2.1 jars. - Added the river 2.2 jars. There are new jars in river which were not part of the jini 2.1 distribution. The new jars only appear in the lib directory (versus the lib-ext and lib-dl directories). The new jars are: asm-3.2.jar, asm-common-3.2.jar, checkconfigurationfile.jar, checkser.jar, classdep.jar, computedigest.jar, computehttpmdcodebase.jar, destroy.jar, envcheck.jar, extra.jar, group.jar, jarwrapper.jar, jsk-debug-policy.jar, outrigger-snaplogstore.jar, phoenix-group.jar, phoenix-init.jar, phoenix.jar, preferredlistgen.jar, serviceui.jar, sharedvm.jar. - No changes were required to build.xml or build.properties. - Renamed the license files for apache zookeeper and apache river to clarify their relationship to the jars. I have not renamed the directory structure within the bigdata-jini project as that would any number of problems with the deployment scripts. - Updated the Depends.java file (in the TERMS branch also). Committed revision r5255.
        Hide
        bryanthompson bryanthompson added a comment -

        The cluster install and TCK / NSS test suites work fine with River in both the 1.0.x and TERMS branches. Both are now running through CI.

        Show
        bryanthompson bryanthompson added a comment - The cluster install and TCK / NSS test suites work fine with River in both the 1.0.x and TERMS branches. Both are now running through CI.

          People

          • Assignee:
            bryanthompson bryanthompson
            Reporter:
            bryanthompson bryanthompson
          • Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved: