Rsync

[Paolo Salvatore]

Hi, I’ll try to write here, I need some help with rsync, I replicate all the ifs to the disaster recovery server, but rsync doesn’t preserve the ccsid attribute, Does anyone have a solutions?

e.g

On the source server I have a stream file with ccsid 1208, when rsync replicate it to the remote server the file created in remote have a ccsid of 819.

Many thanks.

Bye

[Jesse G]

Generally speaking, open source software has no knowledge of file CCSID tags, since that is a concept unique to IBM i and z/OS. As a result, when new files are created, they are tagged with the PASE CCSID of the process (often 819).

For what it’s worth, IBM i 7.4 has changed the default PASE CCSID to 1208 to make these types of problems less prevalent.
Assigning to @Kevin Adler in case he has more to add.

[Paolo Salvatore]

Ok, thanks. I’m on V7R3 and for the moment I can’t upgrade to V7R4 because I have some IBM I that can’t miigrate to this release.

‌

[Kevin Adler]

Jesse is correct that other platforms do not have the knowledge of CCSID tags. However, it’s possible we could add something by mapping the CCSID in to an extended attribute and then the attribute could be preserved across any platform that supported extended attributes (though it wouldn’t be used by anything but IBM i).

Currently extended attribute support in PASE is not very good, so this would require additional work outside the scope of the open source team. I do not see this support being added anytime soon, however feel free to open an RFE if you would like to see this supported.