Hi Experts,
We are planning to migrate our entire project from Unix to Windows platform, also, upgrading to version 9.1 from 7.5.
We would really appreciate if anyone could share some tips about platform migration from Unix to Windows.
Tips like what needs to be taken care of before and after migration.
Cross platform migration
Moderators: chulett, rschirm, roy
Cross platform migration
Known is a drop unknown is an ocean
Depending upone what sort of a project you have and what functionality you used, the re-hosting can be simple or a nightmare.
Paths are first on the list. Windows is quite happy with a path "D:\xxx/yyy\xxx.txt" while UNIX doesn't find it palatable at all.
What sort of a PX project do you have? The number of jobs and sequences is more important than their complexity. Do you access lots of text files or mainly tables? Any user-written buildops or external functions (DS/BASIC functions will generally port with no issues)?
Paths are first on the list. Windows is quite happy with a path "D:\xxx/yyy\xxx.txt" while UNIX doesn't find it palatable at all.
What sort of a PX project do you have? The number of jobs and sequences is more important than their complexity. Do you access lots of text files or mainly tables? Any user-written buildops or external functions (DS/BASIC functions will generally port with no issues)?
</a>Thanks for the response.
Ours is fairly complex project . We do access a lot of tables , do have user written BASIC/C++ routines. We have job routines written as well.
Total job count could be anywhere between 700-1000 jobs.
We do not much of hard-coding . Whatever could be parametrized is already done.
So far , we have checked the paths and rest of the environment variables.
Ours is fairly complex project . We do access a lot of tables , do have user written BASIC/C++ routines. We have job routines written as well.
Total job count could be anywhere between 700-1000 jobs.
We do not much of hard-coding . Whatever could be parametrized is already done.
So far , we have checked the paths and rest of the environment variables.
Known is a drop unknown is an ocean
Then it seems you have a good start already. The c++ code will have to be modified, usually only very slightly, to compile and run on Windows. The BASIC stuff won't notice that it has been moved from UNIX to Windows (but will need to be recompiled).
You will most likely be up-and-running very quickly after having done the connectivity, paths and recompiles but most of your effort will be in doing data regression testing and for tracking down those elusive issues which are not known ahead of time.
You will most likely be up-and-running very quickly after having done the connectivity, paths and recompiles but most of your effort will be in doing data regression testing and for tracking down those elusive issues which are not known ahead of time.