| || |
Monitor builds to test - use CI or custom solution
Our build & release dept wants to de-integrate automated testing from being kicked off as part of their build process. They have been writing to a text file (similar to INI or *nix config file format) which indicates what current build is, and QA can check off that file whether a new build came.
One could poll the file for changes or use something like .NET's FileSystemWatcher class (or similar for other platforms).
What I wanted to ask is whether in such a case, it would be best to build own custom monitoring solution, or whether existing Continuous Integration (CI) tools can be used for this job?
One caveat though is that the build is binary, and QA can't really key off source code changes, etc. to kick off a build and test. Build dept makes the build and then QA can test off that if it finds new build from build dept. So we just key off those text files.
Also, anyone have experience with FileSystemWatcher? Wondering whether that works better in practice or polling? The text files is stored over network fileshare via UNC path.
And are there equivalents to .NET FileSystemWatcher class say in Java, Perl, Python? If yes, what is the equivalent called in those platforms?
Thanks in advance.
Re: Monitor builds to test - use CI or custom solution
Use inbuilt functions available for directories and file handling to get
1. The location of the directory
2. Number of files in the directory
3. File name
Manipulate the code to test the version, Rename etc.,