Step 2 – Migrate Dataset Data
If using an external network-attached storage device (NAS), this step is not necessary. It is assumed that the NFS server that was configured during the installation is the same NAS device that the existing 6.x environment is using.
If the application and/or analytics hosts in the 6.x environments are serving as the NFS shares for /data and /localdata, perform the steps below to copy data from the 6.x hosts to the 7.0 application host. This step can be done ahead of time, in cases where there is a large amount of data that needs to be transferred. If done ahead of time, however, an additional rsync will need to be run at the time of the migration in order to migrate any data that was added since the first rsync was performed. This additional rsync should be much faster, as it should only need to transfer the delta between the data transferred previously compared to the updated content.
On the 6.x application host, run the following commands:
rsync -arzP --omit-dir-times /data/brainspace/ <7.0 application host name/ip>:/data/brainspace/
rsync -arzP --omit-dir-times /localdata/brainspace/datasets/ <7.0 application host name/ip>:/localdata/brainspace/datasets/
It is our recommendation that a screen session be used when doing these commands, as they can take a while to run depending on the amount of data being moved. More information about screen can be found here:
https://www.tecmint.com/screen-command-examples-to-manage-linux-terminals/
Once the rsync is finished, it is a good idea to ensure that the files in the destination directories have the correct permissions. To do this, run the following on the 7.0 application host:
chown -R 1604:1604 /data/brainspace
chown -R 1604:1604 /localdata/brainspace
Note
If there are DAT or native files that exist on the 6.x machine outside of /data/brainspace, they will need to be rsync’ed over to the 7.0 application host and made available at the same path that they exist on the 6.x machine.
In addition, if the paths to these files are outside of /data/brainspace or made available by a file share outside of /data/brainspace, the Docker compose configuration for the brains-app service will need to be updated to add a bind mount for that other location.
For example:
brains-app:
volumes:
- /my/dat_files:/my/dat_files