mds: fix the description for inotable testing only options

The description text are mixed for mds_kill_skip_replaying_inotable
and mds_inject_skip_replaying_inotable.

At the same time rename "mds_kill_skip_replaying_inotable", which
is a bit confusing to "mds_kill_after_journal_logs_flushed".

Fixes: https://tracker.ceph.com/issues/61660
Signed-off-by: Xiubo Li <xiubli@redhat.com>

PendingReleaseNotes

@@ -236,6 +236,10 @@
   "mgr/prometheus/x/server_port" will be displayed instead of
   "mgr/prometheus/server_port". This matches the output of the non pretty-print
   formatted version of the command.
+* CEPHFS: MDS config option name "mds_kill_skip_replaying_inotable" is a bit
+  confusing with "mds_inject_skip_replaying_inotable", therefore renaming it to
+  "mds_kill_after_journal_logs_flushed"
+
 
 >=17.2.1
 

doc/cephfs/mds-config-ref.rst

@@ -56,7 +56,7 @@
 .. confval:: mds_kill_link_at
 .. confval:: mds_kill_rename_at
 .. confval:: mds_inject_skip_replaying_inotable
-.. confval:: mds_kill_skip_replaying_inotable
+.. confval:: mds_kill_after_journal_logs_flushed
 .. confval:: mds_wipe_sessions
 .. confval:: mds_wipe_ino_prealloc
 .. confval:: mds_skip_ino

qa/tasks/cephfs/test_misc.py

@@ -632,7 +632,7 @@ def test_alloc_cinode_assert(self):
         status = self.fs.status()
         rank0 = self.fs.get_rank(rank=0, status=status)
 
-        self.fs.mds_asok(['config', 'set', 'mds_kill_skip_replaying_inotable', "true"])
+        self.fs.mds_asok(['config', 'set', 'mds_kill_after_journal_logs_flushed', "true"])
         # This will make the MDS crash, since we only have one MDS in the
         # cluster and without the "wait=False" it will stuck here forever.
         self.mount_a.run_shell(["mkdir", "test_alloc_ino/dir1"], wait=False)

src/common/options/mds.yaml.in

@@ -1117,14 +1117,14 @@ options:
   default: false
   services:
   - mds
-- name: mds_kill_skip_replaying_inotable
+- name: mds_kill_after_journal_logs_flushed
   type: bool
   level: dev
   default: false
   services:
   - mds
-  fmt_desc: Ceph will skip replaying the inotable when replaying the journal, and
-    the premary MDS will crash, while the replacing MDS won't.
+  fmt_desc: The primary MDS will crash just after the mknod/openc journal logs
+    are flushed to the pool.
     (for testing only).
   with_legacy: true
 - name: mds_inject_skip_replaying_inotable
@@ -1133,8 +1133,7 @@ options:
   default: false
   services:
   - mds
-  fmt_desc: Ceph will skip replaying the inotable when replaying the journal, and
-    the premary MDS will crash, while the replacing MDS won't.
+  fmt_desc: MDS will skip replaying the inotable when replaying the journal logs.
     (for testing only).
   with_legacy: true
 #  percentage of MDS modify replies to skip sending the client a trace on [0-1]

src/mds/Server.cc

@@ -4557,7 +4557,7 @@ class C_MDS_openc_finish : public ServerLogContext {
     ceph_assert(r == 0);
 
     // crash current MDS and the replacing MDS will test the journal
-    ceph_assert(!g_conf()->mds_kill_skip_replaying_inotable);
+    ceph_assert(!g_conf()->mds_kill_after_journal_logs_flushed);
 
     dn->pop_projected_linkage();
 
@@ -6874,7 +6874,7 @@ class C_MDS_mknod_finish : public ServerLogContext {
     ceph_assert(r == 0);
 
     // crash current MDS and the replacing MDS will test the journal
-    ceph_assert(!g_conf()->mds_kill_skip_replaying_inotable);
+    ceph_assert(!g_conf()->mds_kill_after_journal_logs_flushed);
 
     // link the inode
     dn->pop_projected_linkage();
@@ -7166,7 +7166,7 @@ void Server::handle_client_symlink(MDRequestRef& mdr)
   mds->balancer->maybe_fragment(dir, false);
 
   // flush the journal as soon as possible
-  if (g_conf()->mds_kill_skip_replaying_inotable) {
+  if (g_conf()->mds_kill_after_journal_logs_flushed) {
     mdlog->flush();
   }
 }