Is Clearex recursive labeling more efficient than element-by-element labeling?
I am using Clearcase for a ~ 3700 file project. When building a release, there are several directories that need to be tagged recursively. The number of files here is ~ 2400. Files in other directories are selectively labeled. I am currently placing files using a script to iterate over the list of files and mark the files one by one. It takes about 20 minutes when connected to Ethernet.
Is it efficient to divide the marking process into two parts, one for recursive marking and one for selective marking? Is recursive marking faster by 1x, 2x, ...?
source to share
ClearCase file-by-file operations are always ... slow!
You need to apply recursive labels if you can (that is, if all files in a given tree require labeling).
We recommend that you also do this operation dynamically to avoid the side effect of the refresh state of the snapshot view (not refreshed or refreshed)
The result will be faster (I don't have an exact estimate, but at least 2x seems to be correct).
Warning. Your directory you are applying from recursively must be in the correct version (for example, the version selected by the config spec ).
Remember, the point of marking is to identify a coherent set of files (i.e., a set of files that evolve and are labeled as everything). This means that " mklabel -rec
" is always better than putting the shortcut in one file.
Recursive label does not skip a single file, a shortcut placed in files from the list may result in an incomplete set (for example, if the file list on the label is out of date or incomplete)
source to share