From: Matan Breizman <mbreizma@redhat.com>
Date: Wed, 2 Apr 2025 09:30:45 +0000 (+0000)
Subject: crimson/osd/../client_request: add note to can_serve_replica_read
X-Git-Tag: v20.3.0~151^2
X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=480b7c6e19d9061fa279a0255eaab3a532101f42;p=ceph.git

crimson/osd/../client_request: add note to can_serve_replica_read

can_serve_replica_read uses PGLog::has_write_since, when checking for
writes we actually check if any pglog entry belongs to the head object.

The only two pglog entries that are of a clone object are:
1) At creation (pg_log_entry_t::CLONE)
2) At trimming (See remove_or_update)

In both cases, the there would be another pg log entry of the head.

---

Add assertion in prepare_head_update to assert that the above is true.
The obc passed to prepare_head_update (by OpsExecuter) could also be
a clone object (after being resolved). However, write operations should
only occur to head - so let's verify that.

Signed-off-by: Matan Breizman <mbreizma@redhat.com>
---

diff --git a/src/crimson/osd/ops_executer.cc b/src/crimson/osd/ops_executer.cc
index 991c16d7aae..e289a87733a 100644
--- a/src/crimson/osd/ops_executer.cc
+++ b/src/crimson/osd/ops_executer.cc
@@ -903,6 +903,7 @@ pg_log_entry_t OpsExecuter::prepare_head_update(
 {
   LOG_PREFIX(OpsExecuter::prepare_head_update);
   assert(obc->obs.oi.soid.snap >= CEPH_MAXSNAP);
+  assert(obc->obs.oi.soid.is_head());
 
   update_clone_overlap();
   if (cloning_ctx) {
diff --git a/src/crimson/osd/osd_operations/client_request.cc b/src/crimson/osd/osd_operations/client_request.cc
index f7a6b81d374..e97525f9ad1 100644
--- a/src/crimson/osd/osd_operations/client_request.cc
+++ b/src/crimson/osd/osd_operations/client_request.cc
@@ -205,6 +205,8 @@ ClientRequest::interruptible_future<> ClientRequest::with_pg_process_interruptib
       co_await reply_op_error(pgref, -EAGAIN);
       co_return;
     } else if (!pg.get_peering_state().can_serve_replica_read(m->get_hobj())) {
+      // Note: can_serve_replica_read checks for writes on the head object
+      //       as writes can only occur to head.
       DEBUGDPP("{}.{}: unstable write on replica, bouncing to primary",
 	       pg, *this, this_instance_id);
       pg.get_perf_logger().inc(l_osd_replica_read_redirect_conflict);