client: do not send metrics until the MDS rank is ready

In some cases when there are a lot of clients and these clients
have a lots of known requests need to replay too, the metrics
requests will be dropped by the MDS because the MDS is still in
the clientreplay state, and also the useless metric requests will
slow down MDS.

Fixes: https://tracker.ceph.com/issues/61523
Signed-off-by: Xiubo Li <xiubli@redhat.com>

diff --git a/src/client/Client.cc b/src/client/Client.cc
index a456ea2b81fe64d3c388c510d3da2fb95f73e679..89b74065b65a7e70ae64c32ae4f1081ee3c5aeb3 100644 (file)
--- a/src/client/Client.cc
+++ b/src/client/Client.cc
@@ -6927,6 +6927,13 @@ void Client::collect_and_send_global_metrics() {
   ldout(cct, 20) << __func__ << dendl;
   ceph_assert(ceph_mutex_is_locked_by_me(client_lock));
 
+  /* Do not send the metrics until the MDS rank is ready */
+  if (!mdsmap->is_active((mds_rank_t)0)) {
+    ldout(cct, 5) << __func__ << " MDS rank 0 is not ready yet -- not sending metric"
+                  << dendl;
+    return;
+  }
+
   if (!have_open_session((mds_rank_t)0)) {
     ldout(cct, 5) << __func__ << ": no session with rank=0 -- not sending metric"
                   << dendl;