rgw: add rgw_bucket_sync_status() to bypass manager

RGWBucketSyncStatusManager::init() is doing a lot of extra work that's
not needed to serve the rest api (spawning an http manager thread,
fetching the bucket instance info, etc)

uses RGWShardCollectCR to limit the number of concurrent reads to 16

Signed-off-by: Casey Bodley <cbodley@redhat.com>
(cherry picked from commit 403d139c673180ce00eab420a2fdebdacf366dac)

diff --git a/src/rgw/rgw_data_sync.cc b/src/rgw/rgw_data_sync.cc
index 610ab3a500c6219a82d2cbd013d93b78d0470f68..ec8d93b8789d52de09a95ac5b60450ff95797912 100644 (file)
--- a/src/rgw/rgw_data_sync.cc
+++ b/src/rgw/rgw_data_sync.cc
@@ -2967,6 +2967,61 @@ string RGWBucketSyncStatusManager::status_oid(const string& source_zone,
   return bucket_status_oid_prefix + "." + source_zone + ":" + bs.get_key();
 }
 
+class RGWCollectBucketSyncStatusCR : public RGWShardCollectCR {
+  static constexpr int max_concurrent_shards = 16;
+  RGWRados *const store;
+  RGWDataSyncEnv *const env;
+  const int num_shards;
+  rgw_bucket_shard bs;
+
+  using Vector = std::vector<rgw_bucket_shard_sync_info>;
+  Vector::iterator i, end;
+
+ public:
+  RGWCollectBucketSyncStatusCR(RGWRados *store, RGWDataSyncEnv *env,
+                               int num_shards, const rgw_bucket& bucket,
+                               Vector *status)
+    : RGWShardCollectCR(store->ctx(), max_concurrent_shards),
+      store(store), env(env), num_shards(num_shards),
+      bs(bucket, num_shards > 0 ? 0 : -1), // start at shard 0 or -1
+      i(status->begin()), end(status->end())
+  {}
+
+  bool spawn_next() override {
+    if (i == end) {
+      return false;
+    }
+    spawn(new RGWReadBucketSyncStatusCoroutine(env, bs, &*i), false);
+    ++i;
+    ++bs.shard_id;
+    return true;
+  }
+};
+
+int rgw_bucket_sync_status(RGWRados *store, const std::string& source_zone,
+                           const rgw_bucket& bucket,
+                           std::vector<rgw_bucket_shard_sync_info> *status)
+{
+  // read the bucket instance info for num_shards
+  RGWObjectCtx ctx(store);
+  RGWBucketInfo info;
+  int ret = store->get_bucket_instance_info(ctx, bucket, info, nullptr, nullptr);
+  if (ret < 0) {
+    return ret;
+  }
+  status->clear();
+  status->resize(std::max<size_t>(1, info.num_shards));
+
+  RGWDataSyncEnv env;
+  RGWSyncModuleInstanceRef module; // null sync module
+  env.init(store->ctx(), store, nullptr, store->get_async_rados(),
+           nullptr, nullptr, nullptr, source_zone, module);
+
+  RGWCoroutinesManager crs(store->ctx(), store->get_cr_registry());
+  return crs.run(new RGWCollectBucketSyncStatusCR(store, &env, info.num_shards,
+                                                  bucket, status));
+}
+
 
 // TODO: move into rgw_data_sync_trim.cc
 #undef dout_prefix

diff --git a/src/rgw/rgw_data_sync.h b/src/rgw/rgw_data_sync.h
index 56f2995906404a5316ca15ddb6f7c8684f30d612..debfe8d2a51cdb62da2871a4b0a76de0df773d5a 100644 (file)
--- a/src/rgw/rgw_data_sync.h
+++ b/src/rgw/rgw_data_sync.h
@@ -526,6 +526,11 @@ public:
   int run();
 };
 
+/// read the sync status of all bucket shards from the given source zone
+int rgw_bucket_sync_status(RGWRados *store, const std::string& source_zone,
+                           const rgw_bucket& bucket,
+                           std::vector<rgw_bucket_shard_sync_info> *status);
+
 class RGWDefaultSyncModule : public RGWSyncModule {
 public:
   RGWDefaultSyncModule() {}