rgw: reduce default thread pool size

Fixes: https://tracker.ceph.com/issues/65656
Signed-off-by: Mark Kogan <mkogan@redhat.com>

diff --git a/PendingReleaseNotes b/PendingReleaseNotes
index 348fa282cc6e9bd9553fcf937da06abf2620da40..b3dc29c681bd7ce4a181abfa66c6784e27b14adc 100644 (file)
--- a/PendingReleaseNotes
+++ b/PendingReleaseNotes
@@ -59,6 +59,10 @@
 * RGW: Fixed bucket notification events so the 'x_amz_request_id' in NotificationEvent now matches the 'x_amz_request_id' returned by the corresponding S3 operation.
 * RGW: The rgw_gc_max_deferred and rgw_gc_max_deferred_entries_size options have been removed, as they did not do anything.
   - Updated the default for rgw_gc_max_queue_size to account for the extra space from removing rgw_gc_max_deferred_entries_size.
+* RGW: The default value of ``rgw_thread_pool_size`` has been reduced from 512 to 128.
+  Benchmarks showed that the lower thread count improves throughput and reduces latency
+  on NVMe and HDD based environments. Users with specific workload requirements can
+  tune this value; see also ``rgw_max_concurrent_requests``.
 
 
 * DASHBOARD: A new Overview landing page provides an at-a-glance cluster summary

diff --git a/src/common/options/rgw.yaml.in b/src/common/options/rgw.yaml.in
index 5f321fcaff6b5b1318a413efa8e251d0565f3200..b9f78b95ddc46a874d2380e8c881d44e6bfe3470 100644 (file)
--- a/src/common/options/rgw.yaml.in
+++ b/src/common/options/rgw.yaml.in
@@ -1282,9 +1282,11 @@ options:
     is, RGW will be able to deal with more concurrent requests at the cost of more
     resource utilization.
   fmt_desc: The size of the thread pool.
-  default: 512
+  default: 128
   services:
   - rgw
+  see_also:
+  - rgw_max_concurrent_requests
   with_legacy: true
 - name: rgw_num_control_oids
   type: int
@@ -3884,6 +3886,7 @@ options:
   - rgw
   see_also:
   - rgw_frontends
+  - rgw_thread_pool_size
 - name: rgw_scheduler_type
   type: str
   level: advanced