Testing concurrent reads and writes#83
Conversation
|
|
||
| let tbl = tbl index_name | ||
|
|
||
| let w = Index.v ~fresh:false ~log_size index_name |
There was a problem hiding this comment.
I'd prefer constructing this inside the test (with a subscoped directory name), along with the pipes below, and passing them explicitly to the helper functions that need them.
These sorts of global test instances seem to lead to a mess 🙂
There was a problem hiding this comment.
Done, but for now, it still depends on the index being populated by the tbl in common.
index_name is defined above and subscoped to _tests. Is this what you meant?
There was a problem hiding this comment.
If this test uses index_name to construct it's instance-under-test, future tests will also test the same instance on disk. I was suggesting doing something like this to make sure each test uses a unique name.
It's not a big deal, however, since my PR will clean up these sorts of details.
There was a problem hiding this comment.
And w.r.t. the table being populated by Common: I'm also working on that 🙂
icristescu
force-pushed
the
co_test
branch
3 times, most recently
from
b1d0f37 to
ca85d57
Compare
icristescu
force-pushed
the
co_test
branch
3 times, most recently
from
7e915e1 to
ae7a216
Compare
|
Merged the concurrent tests in a single file in PR #66. |
2 participants
No description provided.