diff options
| author | Matthew Sackman <matthew@rabbitmq.com> | 2010-07-07 14:09:47 +0100 |
|---|---|---|
| committer | Matthew Sackman <matthew@rabbitmq.com> | 2010-07-07 14:09:47 +0100 |
| commit | 6978ff64d9d952603f5ac3859d65a332eaa4a876 (patch) | |
| tree | 402c2fb674719f388fa791a88e964622446a6cab | |
| parent | 7f2f52216203fb6506f70d39df4ff79d8695b38e (diff) | |
| download | rabbitmq-server-git-6978ff64d9d952603f5ac3859d65a332eaa4a876.tar.gz | |
When pushing β to γ, *always* do the youngest β first as they will stay around the longest and thus have the greatest chance of reducing the number of subsequent βs pushed to γ
| -rw-r--r-- | src/rabbit_variable_queue.erl | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/rabbit_variable_queue.erl b/src/rabbit_variable_queue.erl index a9e7ccc77f..e566544809 100644 --- a/src/rabbit_variable_queue.erl +++ b/src/rabbit_variable_queue.erl @@ -1133,7 +1133,7 @@ limit_ram_index(State = #vqstate { ram_index_count = RamIndexCount }) -> false -> #vqstate { q2 = Q2, q3 = Q3, index_state = IndexState } = State, {Q2a, {Reduction1, IndexState1}} = - limit_ram_index(fun bpqueue:map_fold_filter_l/4, + limit_ram_index(fun bpqueue:map_fold_filter_r/4, Q2, {Reduction, IndexState}), %% TODO: we shouldn't be writing index %% entries for messages that can never end up |