rgw: add an "object reindex" subcommand to radosgw-admin

This subcommand will take a bucket and an object (or a list of objects in a file) and add those objects to the bucket's index. It does so by preparing the bucket index entry update and then allowing the so-called "dir suggest" mechansim, which is triggered by listing the bucket, to completing the bucket index entry. This mechanism is triggered by listing the bucket. Using this mechanism has the advantage of doing this lazily that both allows the reindex operation to run more quickly and distributes the workload over time. This subcommand can help restore lost bucket index entries. Signed-off-by: J. Eric Ivancich <[email protected]>
ceph · Mar 9, 2023 · be0583c · be0583c
1 parent 44e4dfd
commit be0583c
Show file tree

Hide file tree

Showing 4 changed files with 119 additions and 3 deletions.
diff --git a/src/rgw/driver/rados/rgw_rados.cc b/src/rgw/driver/rados/rgw_rados.cc
@@ -3596,6 +3596,34 @@ int RGWRados::rewrite_obj(RGWBucketInfo& dest_bucket_info, const rgw_obj& obj, c
                        attrset, 0, real_time(), NULL, dpp, y);
 }
 
+int RGWRados::reindex_obj(const RGWBucketInfo& bucket_info,
+			  const rgw_obj& obj,
+			  const DoutPrefixProvider* dpp,
+			  optional_yield y)
+{
+  if (bucket_info.versioned()) {
+    ldpp_dout(dpp, 10) << "WARNING: " << __func__ <<
+      ": cannot process versioned bucket \"" <<
+      bucket_info.bucket.get_key() << "\"" <<
+      dendl;
+    return -ENOTSUP;
+  }
+
+  Bucket target(this, bucket_info);
+  RGWRados::Bucket::UpdateIndex update_idx(&target, obj);
+  const std::string* no_write_tag = nullptr;
+
+  int ret = update_idx.prepare(dpp, RGWModifyOp::CLS_RGW_OP_ADD, no_write_tag, y);
+  if (ret < 0) {
+    ldpp_dout(dpp, 0) << "ERROR: " << __func__ <<
+      ": update index prepare for \"" << obj << "\" returned: " <<
+      cpp_strerror(-ret) << dendl;
+    return ret;
+  }
+
+  return 0;
+}
+
 struct obj_time_weight {
   real_time mtime;
   uint32_t zone_short_id;

diff --git a/src/rgw/driver/rados/rgw_rados.h b/src/rgw/driver/rados/rgw_rados.h
@@ -1075,6 +1075,10 @@ class RGWRados
   D3nDataCache* d3n_data_cache{nullptr};
 
   int rewrite_obj(RGWBucketInfo& dest_bucket_info, const rgw_obj& obj, const DoutPrefixProvider *dpp, optional_yield y);
+  int reindex_obj(const RGWBucketInfo& dest_bucket_info,
+		  const rgw_obj& obj,
+		  const DoutPrefixProvider* dpp,
+		  optional_yield y);
 
   int stat_remote_obj(const DoutPrefixProvider *dpp,
                RGWObjectCtx& obj_ctx,

diff --git a/src/rgw/rgw_admin.cc b/src/rgw/rgw_admin.cc
@@ -72,6 +72,7 @@ extern "C" {
 #include "services/svc_zone.h"
 
 #include "driver/rados/rgw_bucket.h"
+#include "driver/rados/rgw_sal_rados.h"
 
 #define dout_context g_ceph_context
 
@@ -171,6 +172,7 @@ void usage()
   cout << "  object stat                stat an object for its metadata\n";
   cout << "  object unlink              unlink object from bucket index\n";
   cout << "  object rewrite             rewrite the specified object\n";
+  cout << "  object reindex             reindex the object(s) indicated by --bucket and either --object or --objects-file\n";
   cout << "  objects expire             run expired objects cleanup\n";
   cout << "  objects expire-stale list  list stale expired objects (caused by reshard)\n";
   cout << "  objects expire-stale rm    remove stale expired objects\n";
@@ -340,6 +342,7 @@ void usage()
   cout << "   --bucket=<bucket>         Specify the bucket name. Also used by the quota command.\n";
   cout << "   --pool=<pool>             Specify the pool name. Also used to scan for leaked rados objects.\n";
   cout << "   --object=<object>         object name\n";
+  cout << "   --objects-file=<file>     file containing a list of object names to process\n";
   cout << "   --object-version=<version>         object version\n";
   cout << "   --date=<date>             date in the format yyyy-mm-dd\n";
   cout << "   --start-date=<date>       start date in the format yyyy-mm-dd\n";
@@ -425,7 +428,7 @@ void usage()
   cout << "   --show-log-sum=<flag>     enable/disable dump of log summation on log show\n";
   cout << "   --skip-zero-entries       log show only dumps entries that don't have zero value\n";
   cout << "                             in one of the numeric field\n";
-  cout << "   --infile=<file>           specify a file to read in when setting data\n";
+  cout << "   --infile=<file>           file to read in when setting data\n";
   cout << "   --categories=<list>       comma separated list of categories, used in usage show\n";
   cout << "   --caps=<caps>             list of caps (e.g., \"usage=read, write; user=read\")\n";
   cout << "   --op-mask=<op-mask>       permission of user's operations (e.g., \"read, write, delete, *\")\n";
@@ -682,6 +685,7 @@ enum class OPT {
   OBJECT_UNLINK,
   OBJECT_STAT,
   OBJECT_REWRITE,
+  OBJECT_REINDEX,
   OBJECTS_EXPIRE,
   OBJECTS_EXPIRE_STALE_LIST,
   OBJECTS_EXPIRE_STALE_RM,
@@ -898,6 +902,7 @@ static SimpleCmd::Commands all_cmds = {
   { "object unlink", OPT::OBJECT_UNLINK },
   { "object stat", OPT::OBJECT_STAT },
   { "object rewrite", OPT::OBJECT_REWRITE },
+  { "object reindex", OPT::OBJECT_REINDEX },
   { "objects expire", OPT::OBJECTS_EXPIRE },
   { "objects expire-stale list", OPT::OBJECTS_EXPIRE_STALE_LIST },
   { "objects expire-stale rm", OPT::OBJECTS_EXPIRE_STALE_RM },
@@ -3402,6 +3407,7 @@ int main(int argc, const char **argv)
   string op_mask_str;
   string quota_scope;
   string ratelimit_scope;
+  std::string objects_file;
   string object_version;
   string placement_id;
   std::optional<string> opt_storage_class;
@@ -3579,6 +3585,8 @@ int main(int argc, const char **argv)
       pool = rgw_pool(pool_name);
     } else if (ceph_argparse_witharg(args, i, &val, "-o", "--object", (char*)NULL)) {
       object = val;
+    } else if (ceph_argparse_witharg(args, i, &val, "--objects-file", (char*)NULL)) {
+      objects_file = val;
     } else if (ceph_argparse_witharg(args, i, &val, "--object-version", (char*)NULL)) {
       object_version = val;
     } else if (ceph_argparse_witharg(args, i, &val, "--client-id", (char*)NULL)) {
@@ -7774,7 +7782,81 @@ int main(int argc, const char **argv)
     } else {
       ldpp_dout(dpp(), 20) << "skipped object" << dendl;
     }
-  }
+  } // OPT::OBJECT_REWRITE
+
+  if (opt_cmd == OPT::OBJECT_REINDEX) {
+    if (bucket_name.empty()) {
+      cerr << "ERROR: --bucket not specified." << std::endl;
+      return EINVAL;
+    }
+    if (object.empty() && objects_file.empty()) {
+      cerr << "ERROR: neither --object nor --objects-file specified." << std::endl;
+      return EINVAL;
+    } else if (!object.empty() && !objects_file.empty()) {
+      cerr << "ERROR: both --object and --objects-file specified and only one is allowed." << std::endl;
+      return EINVAL;
+    } else if (!objects_file.empty() && !object_version.empty()) {
+      cerr << "ERROR: cannot specify --object_version when --objects-file specified." << std::endl;
+      return EINVAL;
+    }
+
+    int ret = init_bucket(user.get(), tenant, bucket_name, bucket_id, &bucket);
+    if (ret < 0) {
+      cerr << "ERROR: could not init bucket: " << cpp_strerror(-ret) <<
+	"." << std::endl;
+      return -ret;
+    }
+
+    rgw::sal::RadosStore* rados_store = dynamic_cast<rgw::sal::RadosStore*>(driver);
+    if (!rados_store) {
+      cerr <<
+	"ERROR: this command can only work when the cluster has a RADOS backing store." <<
+	std::endl;
+      return EPERM;
+    }
+    RGWRados* store = rados_store->getRados();
+
+    auto process = [&](const std::string& p_object, const std::string& p_object_version) -> int {
+      std::unique_ptr<rgw::sal::Object> obj = bucket->get_object(p_object);
+      obj->set_instance(p_object_version);
+      ret = store->reindex_obj(bucket->get_info(), obj->get_obj(), dpp(), null_yield);
+      if (ret < 0) {
+	return ret;
+      }
+      return 0;
+    };
+
+    if (!object.empty()) {
+      ret = process(object, object_version);
+      if (ret < 0) {
+	return -ret;
+      }
+    } else {
+      std::ifstream file;
+      file.open(objects_file);
+      if (!file.is_open()) {
+	std::cerr << "ERROR: unable to open objects-file \"" <<
+	  objects_file << "\"." << std::endl;
+	return ENOENT;
+      }
+
+      std::string obj_name;
+      const std::string empty_version;
+      while (std::getline(file, obj_name)) {
+	ret = process(obj_name, empty_version);
+	if (ret < 0) {
+	  std::cerr << "ERROR: while processing \"" << obj_name <<
+	    "\", received " << cpp_strerror(-ret) << "." << std::endl;
+	  if (!yes_i_really_mean_it) {
+	    std::cerr <<
+	      "NOTE: with *caution* you can use --yes-i-really-mean-it to push through errors and continue processing." <<
+	      std::endl;
+	    return -ret;
+	  }
+	}
+      } // while
+    }
+  } // OPT::OBJECT_REINDEX
 
   if (opt_cmd == OPT::OBJECTS_EXPIRE) {
     if (!static_cast<rgw::sal::RadosStore*>(driver)->getRados()->process_expire_objects(dpp())) {

diff --git a/src/test/cli/radosgw-admin/help.t b/src/test/cli/radosgw-admin/help.t
@@ -42,6 +42,7 @@
     object stat                stat an object for its metadata
     object unlink              unlink object from bucket index
     object rewrite             rewrite the specified object
+    object reindex             reindex the object(s) indicated by --bucket and either --object or --objects-file
     objects expire             run expired objects cleanup
     objects expire-stale list  list stale expired objects (caused by reshard)
     objects expire-stale rm    remove stale expired objects
@@ -211,6 +212,7 @@
      --bucket=<bucket>         Specify the bucket name. Also used by the quota command.
      --pool=<pool>             Specify the pool name. Also used to scan for leaked rados objects.
      --object=<object>         object name
+     --objects-file=<file>     file containing a list of object names to process
      --object-version=<version>         object version
      --date=<date>             date in the format yyyy-mm-dd
      --start-date=<date>       start date in the format yyyy-mm-dd
@@ -296,7 +298,7 @@
      --show-log-sum=<flag>     enable/disable dump of log summation on log show
      --skip-zero-entries       log show only dumps entries that don't have zero value
                                in one of the numeric field
-     --infile=<file>           specify a file to read in when setting data
+     --infile=<file>           file to read in when setting data
      --categories=<list>       comma separated list of categories, used in usage show
      --caps=<caps>             list of caps (e.g., "usage=read, write; user=read")
      --op-mask=<op-mask>       permission of user's operations (e.g., "read, write, delete, *")