Hi,
I read one presentation, which is briefing about Datastage at a Higher level.
In "Compilation Notes", they wrote the following points.
On clustered / GRID environments:
Share the project directory to share Transformers
(Must be identical mount point across servers)
Or
set $APT_COPY_TRANSFORM_OPERATOR on first job run to distribute Transformer objects
Build stages must be shared, or distributed manually
Can anyone please explain the meaning in detail.
Thanks,
Bharath.
Job Compilation Notes
Moderators: chulett, rschirm, roy
It will depend upon how your storage is shared among the cluster/grid.
If your project directories and directories where custom code (build stages/custom operators/parallel C++ routines) is compiled are shared (cluster filesystem SAN or NFS) among the cluster members (using the same directory paths), then all members should see the same images and manual transport is unnecessary.
If your project directories and directories where custom code is compiled are NOT shared among the cluster members, then manual transport is necessary.
Regards,
If your project directories and directories where custom code (build stages/custom operators/parallel C++ routines) is compiled are shared (cluster filesystem SAN or NFS) among the cluster members (using the same directory paths), then all members should see the same images and manual transport is unnecessary.
If your project directories and directories where custom code is compiled are NOT shared among the cluster members, then manual transport is necessary.
Regards,
- james wiles
All generalizations are false, including this one - Mark Twain.
All generalizations are false, including this one - Mark Twain.