File Activity

Before performing this analysis, be sure you have some data loaded from a git repository

Now that we have some data in our mongo instance, lets start analyzing it. A useful place to start it to see which files have been changed the most. Files that change frequently are more likely to contain bugs or be bottlenecks in the development process.

We can use the following aggregation to find frequently changed files:

[
    {
        $match: {
            date: {$gt: ISODate("2020-01-01")},
        },
    },
    {
        $unwind: {
            path: "$files",
        },
    },
    {
        $sortByCount: "$files.filename",
    },
]

We can also use the file-activity subcommand to run this aggregation:

mdb-code-insights --config-file config.yml file-activity --since 2020-01-01

src/third_party/wiredtiger/import.data: 1335
etc/evergreen.yml: 892
src/mongo/db/SConscript: 354
src/mongo/db/repl/replication_coordinator_impl.cpp: 309
src/mongo/db/s/SConscript: 299
src/mongo/db/repl/SConscript: 245
SConstruct: 226
src/third_party/wiredtiger/test/evergreen.yml: 223
src/third_party/wiredtiger/src/include/wiredtiger.in: 220
...

Looking at the output, the first source code file that show up is src/mongo/db/repl/replication_coordinator_impl.cpp with 309 changes. We will use this file for future exploration.

Now that have an active file, we can analyze what other files are tightly coupled to it.