Contains how-to information referenced from other chapters in this guide. |
Refer to Git Tools Download.
Refer to GitHub Documentation
Refer to ssh Keys.
trafdeveloper_fork |
Do the following:
# Move to the directory where you want to install the Trafodion source code. cd mysource # Clone the Trafodion source code git clone git://github.com/apache/trafodion.git # Change to the source-tree directory cd trafodion # Register your fork as a remote branch git remote add <your-github-id>_fork git@github.com:<your-github-id>/trafodion |
# Move to the directory where you want to install the Trafodion source code. cd mysource # Clone the Trafodion source code git clone git://github.com/apache/trafodion.git # Register your fork as a remote branch cd trafodion git remote add trafdeveloper_fork git@github.com:trafdeveloper/trafodion |
Do the following:
# Ensure that you have the latest changes git fetch --all # Checkout source git checkout -b TRAFODION-<jira-number> origin/master |
# Ensure that you have the latest changes git fetch --all # Checkout source git checkout -b TRAFODION-1507 origin/master |
Merge changes from the master branch using the following commands:
git fetch origin ; git merge origin/master |
Do the following:
# Commit changes git commit -a # Dry-run check against main branch git push -n <your-github-id>_fork HEAD # Push changes to your private fork git push <your-github-id>_fork TRAFODION-1507 |
# Commit changes git commit -a # Dry-run check against main branch git push -n trafdeveloper_fork HEAD # Push changes to your private fork git push trafdeveloper_fork TRAFODION-1507 |
Use the GitHub web interface to generate pull requests.
Do the following:
git checkout TRAFODION-<jira-number> |
git checkout TRAFODION-1507 |
You can force Trafodion to abort and core dump on a specific error. Use this method:
If you want to catch errors in processes other than sqlci, edit the $MY_SQROOT/etc/ms.env file and add the following line
ABORT_ON_ERROR=<error-number> |
There is only one error number that will take effect, so if you have two lines of ABORT_ON_ERROR, the last one wins.
If you want to catch errors on all nodes of the cluster, copy the edited file to all nodes of the cluster:
pdsh $MY_NODES $MY_SQROOT/etc/ms.env $MY_SQROOT/etc |
If you want to catch errors in the sqlci program, export this environment variable, then run sqlci:
export ABORT_ON_ERROR=<error-number> sqlci |
This will help you to debug when a query run into a SQL error, and it is hard to figure out which code path it runs into. Note that the changes in ms.env will affect all processes on that node and therefore could disrupt operation of your cluster.