diff options
| author | Keith Bostic <keith@wiredtiger.com> | 2013-03-07 12:23:31 -0500 |
|---|---|---|
| committer | Keith Bostic <keith@wiredtiger.com> | 2013-03-07 12:23:31 -0500 |
| commit | d44e31cc0cf419ae65848ee2522b1e4a35fa5219 (patch) | |
| tree | 245aabb9531bab6e654f941dbd16d44268fb9289 | |
| parent | f00a575005cab6feb3795ff07c2fe2c6451dded4 (diff) | |
| download | mongo-d44e31cc0cf419ae65848ee2522b1e4a35fa5219.tar.gz | |
Corret a comment.
| -rw-r--r-- | src/btree/rec_write.c | 5 |
1 files changed, 2 insertions, 3 deletions
diff --git a/src/btree/rec_write.c b/src/btree/rec_write.c index bb59f6e3d71..5dce1b4b0a9 100644 --- a/src/btree/rec_write.c +++ b/src/btree/rec_write.c @@ -1065,9 +1065,8 @@ __rec_split_init(WT_SESSION_IMPL *session, * split pages, because otherwise we could end up splitting one large * packed page over and over. We don't want to pick the minimum size * either, because that penalizes an application that did a bulk load - * and subsequently inserted a few items into packed pages. Currently, - * I'm using 75%, but I have no empirical evidence that's a good value. - * We should leave this as a tuning variable, but probably undocumented. + * and subsequently inserted a few items into packed pages. Currently + * defaulted to 75%, but I have no empirical evidence that's "correct". * * The maximum page size may be a multiple of the split page size (for * example, there's a maximum page size of 128KB, but because the table |