Profiling showed a lot of the time was spent in two helper functions:
- glob() which scans the filesystem for filenames that match a pattern
- get_tags() which parses the "@tags" section of a jstest
Printing the arguments of these functions showed that a lot of the calls
are duplicates, because many suites use the same glob pattern, and many
suites want to know what tags a given test contains.
Since we already assume the resmokeconfig is not changing out from under
us, it makes sense to cache these. I don't think memory usage is a
concern because resmoke is either short-lived (as in find-suites) or it
runs subprocesses that already use a lot of memory (mongod, or a whole
sharded cluster).
On my machine these two changes brought the time down from around
1m 20s to 10s.
GitOrigin-RevId: f4538e44aaca8c1ae11c38cc48b6ec76d2a85d1b
When splitting a path of the form "a/*/b/**", globstar.iglob() would
previously have attempted to walk "a/*" as a literal directory instead
of expanding the glob pattern into a list of directories and walking
each of them.
Split out the passthrough tests into separate suites. The MongoDB
deployment is started up by resmoke.py so that we can record the
success/failure of each individual test in MCI.
Added support for parallel execution of tests by dispatching to
multiple MongoDB deployments.
Added support for grouping different kinds of tests (e.g. C++ unit
tests, dbtests, and jstests) so that they can be run together. This
allows for customizability in specifying what tests to execute when
changes are made to a particular part of the code.
