> What's the VM/CMS side written in?  BAL?

So far, REXX and some Pipes. The basic problem is to authenticate to the
Bacula server and then feed it a fairly simple structured data stream.
Bacula doesn't care what the contents are as long as the client
understands what to do with it and the metadata Bacula expects is there
(basically faking up a Unix fstat entry putting the images into a
pseudofilesystem tree, eg /dasd/image/xxxxx). 

It's all TCP sockets, and other than the data transfer, it's all text
line protocols. I had been going along the TRACKREAD/TRACKWRITE path to
get the data when I shelved the work, but I suspect that it wouldn't be
horrendously hard to also do file-level stuff if you could figure out a
way to reliably encode the CMS file metadata in a transportable way
(maybe use VMFPLCD to render the file transportable on store and then
restore the same way? Seems clunky, but would work reliably. Dunno.) and
detect the difference between a CMS minidisk and a "raw" disk. 

The basic line protocol stuff is documented in some detail on
Bacula.org, and I suspect a better piper than I could easily do a better
pipes-based implementation. Pipe-think still doesn't come easily to me. 

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to