Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Migration of unmigrated content due to installation of a new plugin

Excerpthiddentrue

Instructions for creating an official OpenJPA release

...

Making an OpenJPA 1.0.x or 1.1.x Release

These instructions guide the release manager through the steps of making an official OpenJPA release.

Note

These instructions are specific to the incubating status of OpenJPA. Once OpenJPA is elevated out of incubation, these instructions will need to be updated.

Prerequisites

Prerequisites

  1. You should read the OpenJPA Release Policy to decide on the name of the new release, based on the content.
  2. You should read the Apache Release FAQYou should have read the Apache Guide To Release Management During Incubation
  3. You must have shell access to people.apache.org
  4. You must have the following utilities installed on your local machine and available in your path:

...

  1. For Windows users, install Cygwin in addition to the above utilities
    • Make sure the Net/openssh and Utils/gnupg packages are installed that come with Cygwin installation.
    • Optional: Putty

...

Tasks that need to be performed for each release

...

...

In the examples below, it is assumed that the

...

release name will be 1.0.1, and that the current checked-in version name is 1.0.1-SNAPSHOT and stored in the branch named 1.0.x.

  • Download the latest rat jar.
  • run . In general it's a good idea to keep rat.output.txt outside of the code tree for OpenJPA so that it isn't included in the distribution.Look through rat.output.txt for any files that are missing headers. The output will look something like this
    Look for !????? files. A complete file can be found here.
    As you can see from the example there are some files which do not have license headers. This will be addressed in a future release. Until that time you'll need to evaluate each file on a case by case basis. If you find source code or files which do not have the appropriate headers you should add the header and commit the changes back to trunk.

    You can now move the jars from the staging directory on people.apache.org to the incubator repository.
    This can be done via the maven-staging-plugin. An early version of this plugin is available in svn.

    Monday, 12:00 1

    Make sure the #One time setup steps have been performed

    2

    Run the Release Audit Tool

     

    Code Block
    nonenone
    
    java -jar ${pathToRat.jar} ${directoryContainingOpenJPA} > rat.output.txt
    
    Code Block
    nonenone
    
    *****************************************************
    Summary
    -------
    Notes: 6
    Binaries: 33
    Archives: 0
    Standards: 1270
    
    Apache Licensed: 1261
    Generated Documents: 0
    
    JavaDocs are generated and so license header is optional
    Generated files do not required license headers
    
    9 Unknown Licenses
    
    *******************************
    
    Archives (+ indicates readable, $ unreadable):
    
    
    *****************************************************
      Files with ASL headers will be marked L
      Binary files (which do not require ASL headers) will be marked B
      Compressed archives will be marked A
      Notices, licenses etc will be marked N
      N     ./README.txt
      ASL   ./pom.xml
     !????? ./rat.output.txt
      ASL   ./openjpa-all/pom.xml
      ASL   ./openjpa-examples/pom.xml
      ASL   ./openjpa-examples/src/main/java/build.xml
      ASL   ./openjpa-examples/src/main/java/META-INF/persistence.xml
    

    3

    Verify that LICENSE.txt contains up to date licenses for any dependencies which are included in our distribution.
    Any jars or source code which is included with the OpenJPA distribution that is not covered by the Apache license must be noted in LICENSE.txt. Two examples of this are the persistence and orm dtds (licensed under the CDDL) and Serp. If any new non Apache dependencies have been introduced they will have to be covered here as well. If any discrepancies are found update LICENSE.txt and commit the changes.

    4

    Update BUILDING.txt, CHANGES.txt and RELEASE-NOTES.html

     

    • BUILDING.txt should be included in the source tarball and contains instructions on how to build OpenJPA. Prior to shipping a release we should ensure that those instructions are accurate.
    • CHANGES.txt contains a text representation of all the changes which have been made since the preceding release. Most of the contents of this file can be generated through JIRA's release notes mechanism here
    • RELEASE-NOTES.html contains general information on the OpenJPA project as well as an html version of the changes since the preceding version. The html change log may also be generated via JIRA.
      Changes made to any of these files should also be made to trunk.

    5

    If releasing from the trunk, first make a branch from which to make the release. Releasing from a branch will allow any cosmetic changes that need to be made for the release to be approved to be done without preventing other more disruptive advances in the trunk from potentially causing problems with the release. A branch can be made by running:

     

    Code Block
    nonenone
    
    svn copy -m "OpenJPA Release 0.9.6-incubating"
      https://svn.apache.org/repos/asf/incubator/openjpa/trunk
      https://svn.apache.org/repos/asf/incubator/openjpa/branches/0.9.6-incubating
    

    6

    Update the POMs to remove "-SNAPSHOT" from the version. If you have perl installed, you can easily do it with a single command:

     

    Code Block
    nonenone
    
    perl -pi -e 's;<version>0.9.6-incubating-SNAPSHOT</version>;<version>0.9.6-incubating</version>;g' pom.xml */pom.xml
    

    7

    Commit the POM changes

     

    Code Block
    nonenone
    
    svn commit -m "Updated to version 0.9.6-incubating"
    

    8

    Perform the build with documentation and install it in the local repository (this step is required because there is a bug in Maven's javadoc generation aggregated between multiple modules):

     

    Code Block
    nonenone
    
    mvn clean install -Pdocbook-profile,sign-release
    

     

    Note

    This operation will also sign the release files with the gpg utility using the <username>@apache.org key. If your code signing key is under a different address, specify it by appending the following argument to the command above: -Dopenjpa.release.keyAlias=mysigningalias@somedomain.org

    9

    Verify the signatures:

     

    Code Block
    nonenone
    
    gpg --multifile --verify openjpa-project/target/assembly/*.asc
    

    10

    Now actually build the javadocs and perform the deploy upload using the release profile from settings.xml. This will deploy to a staging directory on people.apache.org instead of deploying to the official incubating repository.

     

    Code Block
    nonenone
    
    mvn verify deploy -Pjavadoc-profile,sign-release,release
    

    11

    Tag the branch with the release number:

     

    Code Block
    nonenone
    
    svn copy -m "OpenJPA Release 0.9.6-incubating"
      https://svn.apache.org/repos/asf/incubator/openjpa/branches/0.9.6-incubating
      https://svn.apache.org/repos/asf/incubator/openjpa/tags/0.9.6-incubating
    

    12

    Update the pom.xml files to the subsequent version with the -SNAPSHOT suffix:

     

    Code Block
    nonenone
    
    perl -pi -e "s;<version>0.9.6-incubating</version>;<version>0.9.7-incubating-SNAPSHOT</version>;g" pom.xml */pom.xml
    

    13

    Commit the POM changes

     

    Code Block
    nonenone
    
    svn commit -m "Updated to version 0.9.7-incubating-SNAPSHOT"
    

    14

    Start a vote for the release on the open-jpa-dev@incubator.apache.org mailing list. Votes made by committers and members of the OpenJPA project are considered binding for this vote. For an example of the mail, see this archived vote

    15

    If the vote is successful after 3 days tally the votes in an email to open-jpa-dev@incubator.apache.org  then mail general@incubator.apache.org starting a vote for authorization to release. For this vote only the members of the IPMC are considered binding. A list of the IPMC members can be found here.

    16

    Once that vote is successful, update the http://cwiki.apache.org/openjpa/downloads.html page with information about the download

    17

    The documentation links at http://cwiki.apache.org/openjpa/documentation.html also needs to be updated. You can do this by checking out a copy of the OpenJPA static site documents into a scratch directory, unpacking the OpenJPA binary into the docs section, adding the new docs, and then committing and updating the docs on the server-side:

     

    Code Block
    nonenone
    
    mkdir /tmp/openjpa-docs/
    cd /tmp/openjpa-docs/
    jar xvf OPENJPA_DIR/openjpa-project-0.9.6-incubating-binary.zip
    
    mkdir /tmp/openjpa-site/
    cd /tmp/openjpa-site/
    svn co https://svn.apache.org/repos/asf/incubator/openjpa/site/docs
    
    cd docs/docs/
    mv /tmp/openjpa-docs/openjpa-0.9.6-incubating/docs/ openjpa-0.9.6-incubating
    rm latest
    ln -s openjpa-0.9.6-incubating latest
    
    svn add openjpa-0.9.6-incubating
    svn commit -m "Added documentation for openjpa-0.9.6-incubating release" .
    
    ssh people.apache.org
    svn co http://svn.apache.org/repos/asf/incubator/openjpa/site/docs /www/incubator.apache.org/openjpa
    

     

    The static site will take a little while to synchronize, but eventually you should see the new documentation version at http://incubator.apache.org/openjpa/docs/. Once you see that, add a link to the new documentation release at http://cwiki.apache.org/openjpa/documentation.html

    18

    Code Block
    nonenone
     svn co http://svn.apache.org/repos/asf/maven/plugins/trunk/maven-stage-plugin/ maven-stage-plugin 
    cd maven-stage-plugin
    mvn clean install
    mvn stage:copy -Dsource="http://people.apache.org/~mikedd/public_html/staging-repository/" \
       -Dtarget="scp://people.apache.org/www/people.apache.org/repo/m2-incubating-repository/" \
       -Dversion=0.9.7-incubating 
    ## This last step may take some time. 

    19

    Update the JIRA versions page to mark the version as 'released', and set the date to the date that the release was approved.

    One time setup

    These setup steps only need to be performed on a particular machine once.

    Info

    Developers using Linux workstations can skip over the references to PuTTY and Cygwin

    Create and install a SSH key

    1

    Install PuTTY

    2

    Use PuttyGen to create a SSH key (see Putty help for details)

    3

    Use PuTTY to ssh to people.apache.org

    4

    Create a ~/.ssh folder

    5

    pscp your SSH public key to ~/authorized_keys

    6

    ssh to p.a.o

    7

    Create a ~\.ssh folder and move authorized_keys there

    8

    Configure putty to use your private key and save the session

    Create a PGP key

    1

    Install cgywin, including utils/gpg

    2

    Generate a key with $ gpg --gen-key

    3

    Backup your cygwin home directory to another media

    4

    Add your key to https://svn.apache.org/repos/asf/incubator/openjpa/site/docs/KEYS

    Update Maven settings for our servers

    ...

    1

    ...

    Create a settings.xml under .m2 (in your Document and Settings folder in Windows)

    ...

     

    ...

    
    <settings xmlns="http://maven.apache.org/POM/4.0.0"
              xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
              xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/settings-1.0.0.xsd">
        <servers>
           <server>
              <id>people.apache.org</id>
              <username>$USERNAME</username>
              <privateKey>$PATH_TO_PRIVATE_KEY</privateKey>
              <directoryPermissions>775</directoryPermissions>
              <filePermissions>644</filePermissions>
           </server>
        </servers>    
        <profiles>
            <profile>
                <id>release</id>
                <properties>
                    <deploy.altRepository>people.apache.org::default::scp://people.apache.org/home/${APACHE_USER_ID}/public_html/staging-repository</deploy.altRepository>
                </properties>
            </profile>
        </profiles>
    </settings>
    

    Monday, 12:01

    Create a sub-branch off of the parent branch from which to make the release. Releasing from a branch will allow any cosmetic changes that need to be made for the release to be approved to be done without preventing other more disruptive advances in the trunk from potentially causing problems with the release. A branch can be made by running:

    none svn copy -m "OpenJPA Release 1.0.1 branch" \ https://svn.apache.org/repos/asf/openjpa/branches/1.0.x \ https://svn.apache.org/repos/asf/openjpa/branches/1.0.1

    Monday, 12:02

    Check out a clean branch from which to build the release:

    none svn checkout https://svn.apache.org/repos/asf/openjpa/branches/1.0.1 cd 1.0.1


    Make sure there is no space characters in the path to the build root subdirectory, i.e. "c:\OpenJPA 1.0.1 Release\build". See "#Space Character in Build Root Path"

    Monday, 12:03

    Update all the POMs to remove "-SNAPSHOT" suffix from the version. If you have perl installed, you can easily do it with a single command:

    none perl -pi -e "s;<version>1.0.1-SNAPSHOT</version>;<version>1.0.1</version>;g" \ pom.xml */pom.xml */*/pom.xml

    Update the <scm.dir> property in the top level pom.xml to the 1.0.1 release. \
    Update the <developer> element in the top level pom.xml if new committers are added or removed. This applies to the 1.0.x branch only.

    Monday, 12:10

    Verify that LICENSE.txt contains up to date licenses for any dependencies which are included in our distribution.
    Any jars or source code which is included with the OpenJPA distribution that is not covered by the Apache license must be noted in LICENSE.txt. Two examples of this are the persistence and orm dtds (licensed under the CDDL) and Serp. If any new non Apache dependencies have been introduced they will have to be covered here as well. If any discrepancies are found update LICENSE.txt and commit the changes.

    Monday, 12:15

    Update BUILDING.txt, CHANGES.txt and RELEASE-NOTES.html

    • BUILDING.txt should be included in the source tarball and contains instructions on how to build OpenJPA. Prior to shipping a release we should ensure that those instructions are accurate.
    • CHANGES.txt contains a text representation of all the changes which have been made since the preceding release. Most of the contents of this file can be generated through JIRA's release notes mechanism here
    • RELEASE-NOTES.html contains general information on the OpenJPA project as well as an html version of the changes since the preceding version. The html change log may also be generated via JIRA.

    Monday, 12:20

    Commit the POM changes

    none svn commit -m "Updated to version 1.0.1 for the release"

    Monday, 12:21

    Perform an initial build and install it in the local repository (this step is only required because of an open bug with Maven's javadoc plugin):

    none mvn clean install -Dtest=false

    Monday, 12:23

    Now build the release locally, which will build and test, run the Apache Release Audit Tool to verify license headers, generate the javadoc and docbook PDF and HTML, run through the JPA TCK, build the source and binariy assemblies in target/site/downloads/, and sign the release files.

    none export MAVEN_OPTS=-Xmx1000m mvn --batch-mode deploy site \ -Ptck-profile,examples-profile,license-verify-profile,javadoc-profile,docbook-profile,sign-release \ -Djava14.jar=${JAVA_HOME}/../../1.4/Classes/classes.jar \ -Dtck.zip=${HOME}/.m2/privaterepos/jpa-1_0b-tck.zip


    This operation will also sign the release files with the gpg utility using the <username>@apache.org key. If your code signing key is under a different address, specify it by appending the following argument to the command above: -Dopenjpa.release.keyAlias=mysigningalias@somedomain.org


    The java14.jar path should be changed to the local install path for the JDK 1.4 rt.jar so that JDK 1.4 verification can take place.


    The tck.zip path should be changed to the local install path to the JPA TCK so that the TCK can be run against the release. If you are not licensed to access the TCK, remove the tck-profile profile and -Dtck.zip property from the command. Ask someone who has access to run the TCK for you.


    The examples-profile has problem running automatically in this maven build task. You will need to run the examples manually. Run the maven command without the examples-profile, then perform the following steps to run the example:

    mkdir openjpa-integration\examples\target\examples unzip target\site\downloads\apache-openjpa-1.0.1-binary.zip -d openjpa-integration\examples\target\examples cd openjpa-integration\examples\target\examples\apache-openjpa-1.0.1\examples\hellojpa ant cd ..\relations ant cd ..\reversemapping ant

    Monday, 13:00

    Verify the signatures:

    none gpg --multifile --verify target/site/downloads/*.asc

    Monday, 13:01

    Upload the release candidate to http://openjpa.apache.org/builds/1.0.1/downloads/:

    none mvn site:deploy

    Bear in mind that uploads to people.apache.org/www/openjpa.apache.org/builds/ are not visible at http://openjpa.apache.org/builds until after the hourly synchronization has taken place, as described at http://www.apache.org/dev/project-site.html.

    Monday, 15:00

    Start a vote for the release on the dev@openjpa.apache.org mailing list. Votes made by committers and members of the OpenJPA project are considered binding for this vote. For an example of the mail, see this archived 1.0.0 vote

    Tuesday, Wednesday

    While waiting for the vote to complete, perform whatever manual review and testing on the release you deem appropriate.

    Thursday, 15:00

    If the vote is successful after 3 days tally the votes in an email to dev@openjpa.apache.org, like this one.

    Thursday, 15:10

    Make the OpenJPA assemblies available to the Apache mirroring system by copying them over on people.apache.org:

    none ssh people.apache.org cp -r /www/openjpa.apache.org/builds/1.0.1/downloads/* /www/www.apache.org/dist/openjpa/1.0.1 chgrp -R openjpa /www/www.apache.org/dist/openjpa/1.0.1 chmod -R g+w /www/www.apache.org/dist/openjpa/1.0.1

    The OpenJPA binary release will be available via the link http://www.apache.org/dyn/closer.cgi/openjpa/1.0.1/apache-openjpa-1.0.1-binary.zip after 24 hours, as per the Apache mirroring information.

    Thursday, 15:15

    Now that the release is locked down, convert the writeable 1.0.1 branch to a (du jure) read-only tag:

    none svn mv -m "OpenJPA Release 1.0.1 tag" \ https://svn.apache.org/repos/asf/openjpa/branches/1.0.1 \ https://svn.apache.org/repos/asf/openjpa/tags/1.0.1

    Thursday, 15:16

    Check out the parent branch and bump up the release number to be the next snapshot version:

    none svn checkout https://svn.apache.org/repos/asf/openjpa/branches/1.0.x cd 1.0.x perl -pi -e "s;<version>1.0.1-SNAPSHOT</version>;<version>1.0.2-SNAPSHOT</version>;g" \ pom.xml */pom.xml */*/pom.xml svn commit -m "Updating version in branch to 1.0.2-SNAPSHOT"

    Thursday, 15:20

    Update the http://cwiki.apache.org/openjpa/downloads.html page with links to the download mirrors, using the existing entries as templates.

    All artifacts (apache-openjpa-1.0.1-binary.zip and apache-openjpa-1.0.1-source.zip) must link to the mirrors, but signatures (apache-openjpa-1.0.1-binary.zip.asc and apache-openjpa-1.0.1-source.zip.asc) must not link to mirrors.

    Thursday, 15:30

    The documentation on the server-side must manually be extracted on people.apache.org, and the links at http://openjpa.apache.org/documentation.html need to be updated with the new versions and the "latest" documentation symbolic links need to be updated:

    none cd /www/openjpa.apache.org/builds/1.0.1/ unzip downloads/apache-openjpa-1.0.1-binary.zip rm /www/openjpa.apache.org/docs/latest ln -fvs ../builds/1.0.1/apache-openjpa-1.0.1/docs/ /www/openjpa.apache.org/docs/latest


    Then point the links on http://openjpa.apache.org/documentation.html to:

    Thursday, 15:45

    Merge the staged maven2 repository jars to with the official repository to rsync. This is currently a difficult process that requires the maven-staging-plugin, which much be built from source. The process looks something like this (executed in the root of your local OpenJPA checkout):

    none svn co http://svn.apache.org/repos/asf/maven/plugins/trunk/maven-stage-plugin/ \ /tmp/maven-stage-plugin mvn -f /tmp/maven-stage-plugin/pom.xml clean install for i in $(find openjpa-* -name m2-repository | egrep -v "openjpa-project|openjpa-integration|openjpa-examples"); do cd $i mvn stage:copy -Dsource=file://. \ -Dtarget=scp://people.apache.org:/www/people.apache.org/repo/m2-ibiblio-rsync-repository \ -Dversion=1.0.1 cd - done

    This process requires Maven 2.0.5. It is currently quite delicate and error-prone. Once the maven-stage-plugin is released, it should be possible to make it more automated.

    Window/Cygwin user: See #Merging local repository to remote repository in Cygwin/Windows

    Linux users: You may need to change the '-Dsource=' option to read '-Dsource=file:{}'. The file://{} syntax may result in errors creating the wagon file with errors like: "Repository path /openjpa-jdbc-5/target/site/m2-repository does not exist, and cannot be created."

    Thursday, 16:00

    Update the JIRA versions page to mark the version as "released", and set the date to the date that the release was approved. You may also need to make a new release entry for the subsequent release.

    Friday, 16:00

    After the mirrors have had time to update (24 hours to be on the safe side), make a news announcement on the OpenJPA wiki.

    Once the news item is made, it won't show up on the front page unless you make some minor edit to the containing page (e.g., adding a newline somewhere).

    Friday, 16:05

    Make an announcement about the release on the users@openjpa.apache.org list (and, for major releases, on the announce@apache.org list as per the Apache Announcement Mailing Lists page). The announcement might look something like this.

    Friday, 16:10

    Make an announcement for the freshmeat.net OpenJPA project (optional)

    Friday, 17:00

    Have a beer and enjoy your weekend while the world's grateful programmers revel in yet another high-quality release of Apache OpenJPA!

    One time setup

    These setup steps only need to be performed on a particular machine once.

    ...

    Developers using Linux workstations can skip over the references to PuTTY and Cygwin

    ...

    Create and install a SSH key

    1

    Install PuTTY

    2a

    Use ssh-keygen to create a SSH key.

    See Authenticating By Public Key (OpenSSH) for a good description on why and how to perform this task.

    2b

    In Windows platform, use PuttyGen to create a SSH key (see Putty help for details).

    • Use "SSH-2 DSA" key type and 1024-bit key size.
    • Copy the content of the "Public key for pasting...." and save it to a file named authorized_keys for later use.
    • The private key saved by PuTTYGen can only be used in Putty configuration.

    3

    pscp your SSH public key authorized_keys saved in last step to ~/authorized_keys

    4

    Use PuTTY to login to people.apache.org

    5

    Create a ~\.ssh folder and change its file mode to 700.

    6

    Move or append ~/authorized_keys to ~/.ssh/authorized_keys and change its file mode to 600.

    • Each public key in the authorized_keys spans only one line.
      • For example: "ssh-dss AAAAB3NzaC1kc3MAAA ..... agBmmfZ9uAbSqA== dsa-key-20071107"
    • '#' in the first column is a comment line.

    7

    Configure putty to use your private key and save the session

    Specify your private key in the "Connection -> SSH -> Auth" category in Putty configuration.

    Create a PGP key

    1

    Install cgywin, including Utils/gnupg and Net/openssh packages, or install gpg from http://www.gnupg.org/(en)/download/index.html

    2

    Generate a key-pair with $ gpg --gen-key using default key kind ("DSA and Elgamal") and ELG-E keys size (2048).

    • The generated keys are stored in $HOME/.gnupg or %HOME%\Application Data\gnupg subdirectory.
    • Save the content in this subdirectory to a safe media. This contains your private key used to sign all the OpenJPA release materials.

    3

    Backup your cygwin home directory to another media

    4

    Add your public key to https://svn.apache.org/repos/asf/openjpa/site/docs/KEYS and /www/www.apache.org/dist/openjpa/KEYS. See the commands describe at the beginning of this KEYS file to perform this task. The gpg key-pair is used to sign the published artifacts for the OpenJPA releases.

    5

    Following the instructions in http://people.apache.org/~henkp/trust/ and ask someone in the OpenJPA project to sign your public key.

    6

    Submit your public key to a key server. E.g. http://pgp.surfnet.nl:11371/ or http://pgp.mit.edu/

    Update Maven settings for our servers

    1

    Create a settings.xml under .m2 (in your Document and Settings folder in Windows)

     

    xmlsettings.xmlsolid <settings xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/settings-1.0.0.xsd"> <servers> <server> <id>people.apache.org</id> <username>$USERNAME</username> <privateKey>$PATH_TO_PRIVATE_KEY</privateKey> <directoryPermissions>775</directoryPermissions> <filePermissions>644</filePermissions> </server> </servers> </settings>

    $PATH_TO_PRIVATE_KEY is the path to the private key generated for ssh. E.g. /home/yourLocalUserId/.ssh/id_dsa.

    Expose a copy of known hosts to Maven

    1

    From cygwin, ssh to people.apache.org, save the public key if prompted, and exit

     

    cygwin will save the known hosts to your ~/.ssh folder, but the script cannot access it there (from Windows)

    2

    From cygwin (not Windows) create another .ssh folder at (question)

    3

    Copy the known_hosts file to the new .ssh folder

    Troubleshooting

    Space Character in Build Root Path

    Description

    If there are spaces in the path to the build root subdirectory, the maven task uses to generate the revision number for the org.apache.openjpa.revision.properties yields incorrect data. For example:

    org.apache.openjpa.revision.propertiessolid revision.number=Type 'svnversion --help' for usage. openjpa.version=1.0.1

    Solution

    Rename the path and remove all spaces.

    Merging local repository to remote repository in Cygwin/Windows

    Description

    The "maven-stage-plugin" is very sensitive to the parameters being passed to it, i.e. the source and target URL properties. When this plugin is used under Cygwin, make sure the following practices are used:

    • Use absolute path in the find command's root directory.
      Problem symptom: $ find . -name m2-repository -not -path "*openjpa-project*" \ -exec mvn -f "c:/tmp/maven-stage-plugin/pom.xml" stage:copy \ -Dsource=file://{} -Dtarget=scp://allee8285@people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository \ -Dversion=1.0.1 \; [INFO] Scanning for projects... [INFO] Searching repository for plugin with prefix: 'stage'. ........ [INFO] Downloading file from the source repository: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Error copying repository from file://./openjpa-all/target/site/m2-repository to \ scp://allee8285@people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository
    • Quote and use the drive name in the path.
      Problem symptom: $ find /cygdrive/c/OpenJPA.1.0.1.Release/1.0.1 -name m2-repository -not -path "*openjpa-project*" \ -exec mvn -f /cygwin/c/tmp/maven-stage-plugin/pom.xml stage:copy \ -Dsource=file://{} -Dtarget=scp://allee8285@people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository \ -Dversion=1.0.1 \; [INFO] Scanning for projects... [INFO] Searching repository for plugin with prefix: 'stage'. ........ [INFO] Downloading file from the source repository: [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Error copying repository from file:///cygdrive/c/OpenJPA.1.0.1.Release/1.0.1/openjpa-all/target/site/m2-repository to scp://allee8285@people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository Embedded error: Could not read from file: c:\cygdrive\c\OpenJPA.1.0.1.Release\1.0.1\openjpa-all\target\site\m2-repository \cygdrive\c\OpenJPA.1.0.1.Release\1.0.1\openjpa-all\target\site\m2-repository (Access is denied.)
    • Specify the people.apache.org user id in the target property.
      Problem symptom: $ find "c:/OpenJPA.1.0.1.Release/1.0.1" -name m2-repository -not -path "*openjpa-project*" \ -exec mvn -f "c:/tmp/maven-stage-plugin/pom.xml" stage:copy -Dsource=file://{} -Dtarget=scp://people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository \ -Dversion=1.0.1 \; [INFO] Scanning for projects... [INFO] Searching repository for plugin with prefix: 'stage'. [INFO] ---------------------------------------------------------------------------- ........ [INFO] Downloading file from the source repository: /org/apache/openjpa/openjpa/maven-metadata.xml.sha1 [INFO] Downloading metadata from the target repository. Password:: ********* ........ Password:: ********* [INFO] ------------------------------------------------------------------------ [ERROR] BUILD ERROR [INFO] ------------------------------------------------------------------------ [INFO] Error copying repository from file://c:/OpenJPA.1.0.1.Release/1.0.1/openjpa-all/target/site/m2-repository to \ scp://people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository

    Solution

    As recommended in the descriptions.

    For example:

    find "c:/OpenJPA.1.0.1.Release/1.0.1" -name m2-repository -not -path "*openjpa-project*" \ -exec mvn -f "c:/tmp/maven-stage-plugin/pom.xml" stage:copy \ -Dsource=file://{} \ -Dtarget=scp://allee8285@people.apache.org/www/people.apache.org/repo/m2-ibiblio-rsync-repository \ -Dversion=1.0.1 \;

    Cygwin/Windows File Path

    Description

    For Cygwin/Windows user: file and folder path names using drive identifier (e.g. C:\OpenJPA Release\1.0.1 ) in commands can be expressed as /cygwin/c/OpenJPA Release/1.0.1/. This form of path name specification may have inconsistent and undesirable behaviors.

    Solution

    Consistently use the following naming conventions:

    • Continue to use the Windows form of path name, e.g. C:\a\b\c
    • Use '/' instead of '\' character as file separator, e.g. C:/a/b/c
    • Quote all path name using '"' character, e.g. "C:/a/b/c"
    • Avoid using space characters in path name, e.g. "C:/OpenJPA.Release/1.0.1"

    "Too many unapproved licenses:"

    Description

    Encounter the "Too many unapproved licenses:" message while running the "license-verify-profile" profile in "mvn deploy site..." step. This is caused by extra artifacts in the build tree that the license verification plugin does not recognized. Examples of these artifacts are:

    1. Eclipse control files, .classpath, .project
    2. User created log files

    Solution

    Avoid the followings:

    1. Don't use Eclipse's svn plugin to "Check out" files to a Eclipse project. Simply use the svn command, as described in the release instructions.
    2. Don't create, directly or indirectly, any files under the release build tree.

    Resources

    Expose a copy of known hosts to Maven

    1

    From cygwin, ssh to people.apache.org, save the public key if prompted, and exit

     

    cygwin will save the known hosts to your ~/.ssh folder, but the script cannot access it there (from Windows)

    2

    From cygwin (not Windows) create another .ssh folder at (question)

    3

    Copy the known_hosts file to the new .ssh folder

    Example shell script to perform the steps above

    ...

    
    #!/bin/sh -pve
    # Author: Marc Prud'hommeaux <mprudhom@apache.org>
    # Performs the release steps described at:
    #
    #   http://cwiki.apache.org/openjpa/releasing-openjpa.html
    #
    # It will do the following:
    #
    # 1. Check out a fresh version of openjpa from the branch
    # 2. Update the openjpa pom.xml files to have the new version
    # 3. Commit the pom.xml changes
    # 4. Make the release files
    # 5. Verify the signature
    # 6. Test the examples in the release
    # 7. Perform the deploy
    # 8. Tag the view using "svn copy"
    
    BASEDIR=/tmp/openjpa-staging/
    RELEASEDIR=${BASEDIR}/openjpa
    EXAMPLESDIR=${BASEDIR}/examples
    
    rm -rf ${BASEDIR} || echo Staging directory already deleted
    
    # OLDVERSION=0.9.6-incubating-SNAPSHOT
    # RELEASEVERSION=0.9.6-incubating
    
    OLDVERSION=${1}
    RELEASEVERSION=${2}
    
    shift;
    shift;
    EXTRAARGS=${@}
    
    # example usage:
    # openjpa.mkrelease 0.9.6-incubating-SNAPSHOT 0.9.6-incubating 0.9.7-incubating-SNAPSHOT
    # openjpa.mkrelease 0.9.6-incubating-SNAPSHOT 0.9.6-incubating 0.9.7-incubating-SNAPSHOT \
    # -Dopenjpa.release.keyAlias=somegpgkeyalias@somedomain.org
    
    test ! -z ${RELEASEVERSION} || echo "Usage: ${0} OLDVERSION RELEASEVERSION"
    test ! -z ${RELEASEVERSION}
    
    # svn co https://svn.apache.org/repos/asf/incubator/openjpa/trunk/ ${RELEASEDIR}
    
    # Check out from the branch (note that a branch with the same name
    # as the release version needs to exist)...
    SVNBASE="https://svn.apache.org/repos/asf/incubator/openjpa"
    TRUNK="${SVNBASE}/trunk"
    BRANCH="${SVNBASE}/branches/${RELEASEVERSION}"
    TAG="${SVNBASE}/tags/${RELEASEVERSION}"
    
    svn co ${BRANCH} ${RELEASEDIR} ||
      \ (echo "$0: Branch does not exist. You can create it \
         from the trunk with: svn copy ${TRUNK} ${BRANCH}" && false)
    
    cd ${RELEASEDIR}
    
    grep "<version>${OLDVERSION}</version>" pom.xml || \
      echo "ERROR: version is not the expected version (${OLDVERSION})"
    grep "<version>${OLDVERSION}</version>" pom.xml
    
    perl -pi -e "s;<version>${OLDVERSION}</version>;<version>${RELEASEVERSION}</version>;g" pom.xml */pom.xml
    
    svn commit -m "Updated to version ${RELEASEVERSION}"
    
    # Pre-build: need to do this to get around bugs in aggregate javadocs, as
    # well as making a signature we can test
    mvn clean install -Pdocbook-profile,sign-release "${EXTRAARGS}"
    
    # Verify the signatures
    gpg --multifile --verify openjpa-project/target/assembly/*.asc
    
    # Test the examples to make sure they work
    OLDDIR=`pwd`
    rm -rf ${EXAMPLESDIR} || true
    mkdir -p ${EXAMPLESDIR}
    cd ${EXAMPLESDIR}
    unzip ${RELEASEDIR}/openjpa-project/target/assembly/*-binary.zip
    
    for build in openjpa-*/examples/*/build.xml
        do
        ant -f ${build}
    done
    
    cd ${OLDDIR}
    
    # Now actually build the javadocs and perform the deploy upload
    mvn verify deploy -Pjavadoc-profile,sign-release "${EXTRAARGS}"
    
    # Remove any identical tag
    svn delete -m "Removed old ${RELEASEVERSION} tag for re-tagging" ${TAG} || \
      echo "Tag does not already exist, so does not need to be removed"
    
    # Now tag the view
    svn copy -m "OpenJPA Release ${RELEASEVERSION}" ${BRANCH} ${TAG}
    
    # We don't do this anymore since we release from branches, not from trunk
    # Update to the next version
    # perl -pi -e "s;<version>${RELEASEVERSION}</version>;<version>${NEXTVERSION}</version>;g" pom.xml */pom.xml
    #
    # # Commit the next versions
    # svn commit -m "Updated to version ${NEXTVERSION}"
    

    Resources