ceph/doc/radosgw/dynamicresharding.rst

   1 .. _rgw_dynamic_bucket_index_resharding:
   2
   3 ===================================
   4 RGW Dynamic Bucket Index Resharding
   5 ===================================
   6
   7 .. versionadded:: Luminous
   8
   9 A large bucket index can lead to performance problems. In order
  10 to address this problem we introduced bucket index sharding.
  11 Until Luminous, changing the number of bucket shards (resharding)
  12 needed to be done offline. Starting with Luminous we support
  13 online bucket resharding.
  14
  15 Each bucket index shard can handle its entries efficiently up until
  16 reaching a certain threshold number of entries. If this threshold is
  17 exceeded the system can suffer from performance issues. The dynamic
  18 resharding feature detects this situation and automatically increases
  19 the number of shards used by the bucket index, resulting in a
  20 reduction of the number of entries in each bucket index shard. This
  21 process is transparent to the user. Write I/Os to the target bucket
  22 are blocked and read I/Os are not during resharding process.
  23
  24 By default dynamic bucket index resharding can only increase the
  25 number of bucket index shards to 1999, although this upper-bound is a
  26 configuration parameter (see Configuration below). When
  27 possible, the process chooses a prime number of bucket index shards to
  28 spread the number of bucket index entries across the bucket index
  29 shards more evenly.
  30
  31 The detection process runs in a background process that periodically
  32 scans all the buckets. A bucket that requires resharding is added to
  33 the resharding queue and will be scheduled to be resharded later. The
  34 reshard thread runs in the background and execute the scheduled
  35 resharding tasks, one at a time.
  36
  37 Multisite
  38 =========
  39
  40 Dynamic resharding is not supported in a multisite environment.
  41
  42
  43 Configuration
  44 =============
  45
  46 Enable/Disable dynamic bucket index resharding:
  47
  48 - ``rgw_dynamic_resharding``:  true/false, default: true
  49
  50 Configuration options that control the resharding process:
  51
  52 - ``rgw_max_objs_per_shard``: maximum number of objects per bucket index shard before resharding is triggered, default: 100000 objects
  53
  54 - ``rgw_max_dynamic_shards``: maximum number of shards that dynamic bucket index resharding can increase to, default: 1999
  55
  56 - ``rgw_reshard_bucket_lock_duration``: duration, in seconds, of lock on bucket obj during resharding, default: 360 seconds (i.e., 6 minutes)
  57
  58 - ``rgw_reshard_thread_interval``: maximum time, in seconds, between rounds of resharding queue processing, default: 600 seconds (i.e., 10 minutes)
  59
  60 - ``rgw_reshard_num_logs``: number of shards for the resharding queue, default: 16
  61
  62 Admin commands
  63 ==============
  64
  65 Add a bucket to the resharding queue
  66 ------------------------------------
  67
  68 ::
  69
  70    # radosgw-admin reshard add --bucket <bucket_name> --num-shards <new number of shards>
  71
  72 List resharding queue
  73 ---------------------
  74
  75 ::
  76
  77    # radosgw-admin reshard list
  78
  79 Process tasks on the resharding queue
  80 -------------------------------------
  81
  82 ::
  83
  84    # radosgw-admin reshard process
  85
  86 Bucket resharding status
  87 ------------------------
  88
  89 ::
  90
  91    # radosgw-admin reshard status --bucket <bucket_name>
  92
  93 The output is a json array of 3 objects (reshard_status, new_bucket_instance_id, num_shards) per shard.
  94
  95 For example, the output at different Dynamic Resharding stages is shown below:
  96
  97 ``1. Before resharding occurred:``
  98 ::
  99
 100   [
 101     {
 102         "reshard_status": "not-resharding",
 103         "new_bucket_instance_id": "",
 104         "num_shards": -1
 105     }
 106   ]
 107
 108 ``2. During resharding:``
 109 ::
 110
 111   [
 112     {
 113         "reshard_status": "in-progress",
 114         "new_bucket_instance_id": "1179f470-2ebf-4630-8ec3-c9922da887fd.8652.1",
 115         "num_shards": 2
 116     },
 117     {
 118         "reshard_status": "in-progress",
 119         "new_bucket_instance_id": "1179f470-2ebf-4630-8ec3-c9922da887fd.8652.1",
 120         "num_shards": 2
 121     }
 122   ]
 123
 124 ``3, After resharding completed:``
 125 ::
 126
 127   [
 128     {
 129         "reshard_status": "not-resharding",
 130         "new_bucket_instance_id": "",
 131         "num_shards": -1
 132     },
 133     {
 134         "reshard_status": "not-resharding",
 135         "new_bucket_instance_id": "",
 136         "num_shards": -1
 137     }
 138   ]
 139
 140
 141 Cancel pending bucket resharding
 142 --------------------------------
 143
 144 Note: Ongoing bucket resharding operations cannot be cancelled. ::
 145
 146    # radosgw-admin reshard cancel --bucket <bucket_name>
 147
 148 Manual immediate bucket resharding
 149 ----------------------------------
 150
 151 ::
 152
 153    # radosgw-admin bucket reshard --bucket <bucket_name> --num-shards <new number of shards>
 154
 155 When choosing a number of shards, the administrator should keep a
 156 number of items in mind. Ideally the administrator is aiming for no
 157 more than 100000 entries per shard, now and through some future point
 158 in time.
 159
 160 Additionally, bucket index shards that are prime numbers tend to work
 161 better in evenly distributing bucket index entries across the
 162 shards. For example, 7001 bucket index shards is better than 7000
 163 since the former is prime. A variety of web sites have lists of prime
 164 numbers; search for "list of prime numbers" withy your favorite web
 165 search engine to locate some web sites.
 166
 167 Troubleshooting
 168 ===============
 169
 170 Clusters prior to Luminous 12.2.11 and Mimic 13.2.5 left behind stale bucket
 171 instance entries, which were not automatically cleaned up. The issue also affected
 172 LifeCycle policies, which were not applied to resharded buckets anymore. Both of
 173 these issues can be worked around using a couple of radosgw-admin commands.
 174
 175 Stale instance management
 176 -------------------------
 177
 178 List the stale instances in a cluster that are ready to be cleaned up.
 179
 180 ::
 181
 182    # radosgw-admin reshard stale-instances list
 183
 184 Clean up the stale instances in a cluster. Note: cleanup of these
 185 instances should only be done on a single site cluster.
 186
 187 ::
 188
 189    # radosgw-admin reshard stale-instances rm
 190
 191
 192 Lifecycle fixes
 193 ---------------
 194
 195 For clusters that had resharded instances, it is highly likely that the old
 196 lifecycle processes would have flagged and deleted lifecycle processing as the
 197 bucket instance changed during a reshard. While this is fixed for newer clusters
 198 (from Mimic 13.2.6 and Luminous 12.2.12), older buckets that had lifecycle policies and
 199 that have undergone resharding will have to be manually fixed.
 200
 201 The command to do so is:
 202
 203 ::
 204
 205    # radosgw-admin lc reshard fix --bucket {bucketname}
 206
 207
 208 As a convenience wrapper, if the ``--bucket`` argument is dropped then this
 209 command will try and fix lifecycle policies for all the buckets in the cluster.
 210
 211 Object Expirer fixes
 212 --------------------
 213
 214 Objects subject to Swift object expiration on older clusters may have
 215 been dropped from the log pool and never deleted after the bucket was
 216 resharded. This would happen if their expiration time was before the
 217 cluster was upgraded, but if their expiration was after the upgrade
 218 the objects would be correctly handled. To manage these expire-stale
 219 objects, radosgw-admin provides two subcommands.
 220
 221 Listing:
 222
 223 ::
 224
 225    # radosgw-admin objects expire-stale list --bucket {bucketname}
 226
 227 Displays a list of object names and expiration times in JSON format.
 228
 229 Deleting:
 230
 231 ::
 232
 233    # radosgw-admin objects expire-stale rm --bucket {bucketname}
 234
 235
 236 Initiates deletion of such objects, displaying a list of object names, expiration times, and deletion status in JSON format.