librbd: fix mirror image status summary in a namespace

For the purposes of the summary with image counts, "rbd mirror pool
status" command is supposed to count each image only once.  To this
end, for unidirectional mirroring the status of the receiving site
should be taken while for bidirectional mirroring the statuses should
be combined/reduced.  For example, if mirroring is enabled on a single
image and everything is in order, the summary is expected to be

  image health: OK
  images: 1 total
      1 replaying

on both clusters even though on the primary the local status is
MIRROR_IMAGE_STATUS_STATE_STOPPED and only on the secondary it's
MIRROR_IMAGE_STATUS_STATE_REPLAYING.

Currently this isn't the case for custom namespaces.  In the same
scenario the primary ends up reporting

  image health: OK
  images: 1 total
      1 stopped

based solely on the local status in a namespace.

Fixes: https://tracker.ceph.com/issues/69911
Signed-off-by: Ilya Dryomov <[email protected]>

Author: idryomov

qa/workunits/rbd/rbd_mirror_bootstrap.sh

@@ -35,12 +35,28 @@ done
 rbd --cluster ${CLUSTER1} --pool ${POOL} mirror pool info --format json | jq -e '.peers[0].direction == "tx-only"'
 
 create_image_and_enable_mirror ${CLUSTER1} ${POOL} image1
+create_image_and_enable_mirror ${CLUSTER1} ${POOL}/${NS1} image1
 
 wait_for_image_replay_started ${CLUSTER2} ${POOL} image1
 write_image ${CLUSTER1} ${POOL} image1 100
 wait_for_replay_complete ${CLUSTER2} ${CLUSTER1} ${POOL} ${POOL} image1
 wait_for_replaying_status_in_pool_dir ${CLUSTER2} ${POOL} image1
 
+POOL_STATUS=$(get_pool_status_json ${CLUSTER1} ${POOL})
+jq -e '.summary.states == {"replaying": 1}' <<< ${POOL_STATUS}
+POOL_STATUS=$(get_pool_status_json ${CLUSTER2} ${POOL})
+jq -e '.summary.states == {"replaying": 1}' <<< ${POOL_STATUS}
+
+wait_for_image_replay_started ${CLUSTER2} ${POOL}/${NS1} image1
+write_image ${CLUSTER1} ${POOL}/${NS1} image1 100
+wait_for_replay_complete ${CLUSTER2} ${CLUSTER1} ${POOL}/${NS1} ${POOL}/${NS1} image1
+wait_for_replaying_status_in_pool_dir ${CLUSTER2} ${POOL}/${NS1} image1
+
+POOL_STATUS=$(get_pool_status_json ${CLUSTER1} ${POOL}/${NS1})
+jq -e '.summary.states == {"replaying": 1}' <<< ${POOL_STATUS}
+POOL_STATUS=$(get_pool_status_json ${CLUSTER2} ${POOL}/${NS1})
+jq -e '.summary.states == {"replaying": 1}' <<< ${POOL_STATUS}
+
 testlog "TEST: verify rx-tx direction"
 # both rx-tx peers are added immediately by "rbd mirror pool peer bootstrap import"
 rbd --cluster ${CLUSTER1} --pool ${PARENT_POOL} mirror pool info --format json | jq -e '.peers[0].direction == "rx-tx"'
@@ -52,6 +68,9 @@ create_image ${CLUSTER2} ${PARENT_POOL} image2
 enable_mirror ${CLUSTER1} ${PARENT_POOL} image1
 enable_mirror ${CLUSTER2} ${PARENT_POOL} image2
 
+create_image_and_enable_mirror ${CLUSTER1} ${PARENT_POOL}/${NS1} image1
+create_image_and_enable_mirror ${CLUSTER2} ${PARENT_POOL}/${NS1} image2
+
 wait_for_image_replay_started ${CLUSTER2} ${PARENT_POOL} image1
 write_image ${CLUSTER1} ${PARENT_POOL} image1 100
 wait_for_replay_complete ${CLUSTER2} ${CLUSTER1} ${PARENT_POOL} ${PARENT_POOL} image1
@@ -62,6 +81,26 @@ write_image ${CLUSTER2} ${PARENT_POOL} image2 100
 wait_for_replay_complete ${CLUSTER1} ${CLUSTER2} ${PARENT_POOL} ${PARENT_POOL} image2
 wait_for_replaying_status_in_pool_dir ${CLUSTER1} ${PARENT_POOL} image2
 
+POOL_STATUS=$(get_pool_status_json ${CLUSTER1} ${PARENT_POOL})
+jq -e '.summary.states == {"replaying": 2}' <<< ${POOL_STATUS}
+POOL_STATUS=$(get_pool_status_json ${CLUSTER2} ${PARENT_POOL})
+jq -e '.summary.states == {"replaying": 2}' <<< ${POOL_STATUS}
+
+wait_for_image_replay_started ${CLUSTER2} ${PARENT_POOL}/${NS1} image1
+write_image ${CLUSTER1} ${PARENT_POOL}/${NS1} image1 100
+wait_for_replay_complete ${CLUSTER2} ${CLUSTER1} ${PARENT_POOL}/${NS1} ${PARENT_POOL}/${NS1} image1
+wait_for_replaying_status_in_pool_dir ${CLUSTER2} ${PARENT_POOL}/${NS1} image1
+
+wait_for_image_replay_started ${CLUSTER1} ${PARENT_POOL}/${NS1} image2
+write_image ${CLUSTER2} ${PARENT_POOL}/${NS1} image2 100
+wait_for_replay_complete ${CLUSTER1} ${CLUSTER2} ${PARENT_POOL}/${NS1} ${PARENT_POOL}/${NS1} image2
+wait_for_replaying_status_in_pool_dir ${CLUSTER1} ${PARENT_POOL}/${NS1} image2
+
+POOL_STATUS=$(get_pool_status_json ${CLUSTER1} ${PARENT_POOL}/${NS1})
+jq -e '.summary.states == {"replaying": 2}' <<< ${POOL_STATUS}
+POOL_STATUS=$(get_pool_status_json ${CLUSTER2} ${PARENT_POOL}/${NS1})
+jq -e '.summary.states == {"replaying": 2}' <<< ${POOL_STATUS}
+
 testlog "TEST: pool replayer and callout cleanup when peer is updated"
 test_health_state ${CLUSTER1} ${PARENT_POOL} 'OK'
 test_health_state ${CLUSTER2} ${PARENT_POOL} 'OK'

qa/workunits/rbd/rbd_mirror_helpers.sh

@@ -299,9 +299,11 @@ setup_pools()
 
     rbd --cluster ${cluster} namespace create ${POOL}/${NS1}
     rbd --cluster ${cluster} namespace create ${POOL}/${NS2}
+    rbd --cluster ${cluster} namespace create ${PARENT_POOL}/${NS1}
 
     rbd --cluster ${cluster} mirror pool enable ${POOL}/${NS1} ${MIRROR_POOL_MODE}
     rbd --cluster ${cluster} mirror pool enable ${POOL}/${NS2} image
+    rbd --cluster ${cluster} mirror pool enable ${PARENT_POOL}/${NS1} ${MIRROR_POOL_MODE}
 
     if [ -z ${RBD_MIRROR_MANUAL_PEERS} ]; then
       if [ -z ${RBD_MIRROR_CONFIG_KEY} ]; then

src/librbd/api/Mirror.cc

@@ -1993,8 +1993,11 @@ int Mirror<I>::image_status_summary(librados::IoCtx& io_ctx,
                                     MirrorImageStatusStates *states) {
   CephContext *cct = reinterpret_cast<CephContext *>(io_ctx.cct());
 
+  librados::IoCtx default_ns_io_ctx;
+  default_ns_io_ctx.dup(io_ctx);
+  default_ns_io_ctx.set_namespace("");
   std::vector<cls::rbd::MirrorPeer> mirror_peers;
-  int r = cls_client::mirror_peer_list(&io_ctx, &mirror_peers);
+  int r = cls_client::mirror_peer_list(&default_ns_io_ctx, &mirror_peers);
   if (r < 0 && r != -ENOENT) {
     lderr(cct) << "failed to list mirror peers: " << cpp_strerror(r) << dendl;
     return r;