We added some patches to TEST that we are not going to move forward on and are not sure the best way to get TEST and LIVE back into synch. Do we REMOVE the patches from TEST? Is there a way to make an exact copy of live (data & program files) to copy over top of TEST? Is there a way for me to quickly determine if LIVE and TEST are in synch in regards to the versions of the files? I know that the last patch affected pa100pd and I can certainly look to see if that file version is in synch..but we are going to have to load 8 patches and it would be impossible for me to know if all program files are matched.
any advice/assistance on this subject is greatly appreciated!
My suggestion would be to copy LIVE over to TEST. It's the only way to make sure your in sync (at least to me). But also copy these directories including *src and *lib:
$LAWDIR/prodline/Admin -- log files, especially for patches.
""/backup -- backup of files replaced by patches
""/metadata -- changes to the internal files information etc
Of course backup your TEST, lawson and data, just in case. I've had to this once or twice. It's not bad except that our test server is old and takes a long time to rebuild the indexes.
copying just the source code will not solve all of your problems necessarily either. Many patches, especially some of the big ones, include metadata updates, particularly work file definition changes and library definition changes. If you loaded the metadata from a patch and then overlayed the source code from a pervious version (aka PROD), then you'll very likely run into problems.
Two approaches from this
1) I would follow the procedure to backout the patches. Make sure to complete these backouts in reverse order from which they were applied.
2) If that won't work for some reason, or you've already done them in the wrong order, then follow the usual product line copy procedures available in a sys admin manual from support.lawson.com. Basically:
Dump / load product line meta data
dump / load security
copy all source code, including libraries, etc
system compile
copy database data
finish up by copying work files, print files, etc as needed
If you try option 1 and want to verify, I would go with a script route. I wrote a perl script that reads an input file of programs you want to compare, and then does a diff between each product line, spitting out if the files are the same of different. I'm in a windows environment and have this working across machines since DEV, TEST and PROD are on different boxes. Not sure what type of "net use" equivalent there might be on Unix.