common/options: bluestore 64k min_alloc_size for HDD

Until we understand the performance regression and allocator behavior,
go back to 64k.  This will continue to have a high space overhead for
small objects and EC, but will preserve the current performance levels
for all workloads.

This partially reverts 0ec75c99ddb103c664f66c04450003ed4c407708

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 66797ef19949503e77d52b567962c218ff0d8a77)

diff --git a/src/common/options.cc b/src/common/options.cc
index 78af77415f9422718586c1fb77889e30fd4d91a8..5e401145fa56b3cef40bb02062e3ccac334f2ba2 100644 (file)
--- a/src/common/options.cc
+++ b/src/common/options.cc
@@ -4501,7 +4501,7 @@ std::vector<Option> get_global_options() {
     .set_long_description("A smaller allocation size generally means less data is read and then rewritten when a copy-on-write operation is triggered (e.g., when writing to something that was recently snapshotted).  Similarly, less data is journaled before performing an overwrite (writes smaller than min_alloc_size must first pass through the BlueStore journal).  Larger values of min_alloc_size reduce the amount of metadata required to describe the on-disk layout and reduce overall fragmentation."),
 
     Option("bluestore_min_alloc_size_hdd", Option::TYPE_SIZE, Option::LEVEL_ADVANCED)
-    .set_default(4_K)
+    .set_default(64_K)
     .set_flag(Option::FLAG_CREATE)
     .set_description("Default min_alloc_size value for rotational media")
     .add_see_also("bluestore_min_alloc_size"),