]> git.proxmox.com Git - ceph.git/blob - ceph/doc/radosgw/multisite.rst
import 15.2.4
[ceph.git] / ceph / doc / radosgw / multisite.rst
1 .. _multisite:
2
3 ==========
4 Multi-Site
5 ==========
6
7 .. versionadded:: Jewel
8
9 A single zone configuration typically consists of one zone group containing one
10 zone and one or more `ceph-radosgw` instances where you may load-balance gateway
11 client requests between the instances. In a single zone configuration, typically
12 multiple gateway instances point to a single Ceph storage cluster. However, Kraken
13 supports several multi-site configuration options for the Ceph Object Gateway:
14
15 - **Multi-zone:** A more advanced configuration consists of one zone group and
16 multiple zones, each zone with one or more `ceph-radosgw` instances. Each zone
17 is backed by its own Ceph Storage Cluster. Multiple zones in a zone group
18 provides disaster recovery for the zone group should one of the zones experience
19 a significant failure. In Kraken, each zone is active and may receive write
20 operations. In addition to disaster recovery, multiple active zones may also
21 serve as a foundation for content delivery networks.
22
23 - **Multi-zone-group:** Formerly called 'regions', Ceph Object Gateway can also
24 support multiple zone groups, each zone group with one or more zones. Objects
25 stored to zones in one zone group within the same realm as another zone
26 group will share a global object namespace, ensuring unique object IDs across
27 zone groups and zones.
28
29 - **Multiple Realms:** In Kraken, the Ceph Object Gateway supports the notion
30 of realms, which can be a single zone group or multiple zone groups and
31 a globally unique namespace for the realm. Multiple realms provide the ability
32 to support numerous configurations and namespaces.
33
34 Replicating object data between zones within a zone group looks something
35 like this:
36
37 .. image:: ../images/zone-sync2.png
38 :align: center
39
40 For additional details on setting up a cluster, see `Ceph Object Gateway for
41 Production <https://access.redhat.com/documentation/en-us/red_hat_ceph_storage/3/html/ceph_object_gateway_for_production/index/>`__.
42
43 Functional Changes from Infernalis
44 ==================================
45
46 In Kraken, you can configure each Ceph Object Gateway to
47 work in an active-active zone configuration, allowing for writes to
48 non-master zones.
49
50 The multi-site configuration is stored within a container called a
51 "realm." The realm stores zone groups, zones, and a time "period" with
52 multiple epochs for tracking changes to the configuration. In Kraken,
53 the ``ceph-radosgw`` daemons handle the synchronization,
54 eliminating the need for a separate synchronization agent. Additionally,
55 the new approach to synchronization allows the Ceph Object Gateway to
56 operate with an "active-active" configuration instead of
57 "active-passive".
58
59 Requirements and Assumptions
60 ============================
61
62 A multi-site configuration requires at least two Ceph storage clusters,
63 preferably given a distinct cluster name. At least two Ceph object
64 gateway instances, one for each Ceph storage cluster.
65
66 This guide assumes at least two Ceph storage clusters are in geographically
67 separate locations; however, the configuration can work on the same
68 site. This guide also assumes two Ceph object gateway servers named
69 ``rgw1`` and ``rgw2``.
70
71 .. important:: Running a single Ceph storage cluster is NOT recommended unless you have
72 low latency WAN connections.
73
74 A multi-site configuration requires a master zone group and a master
75 zone. Additionally, each zone group requires a master zone. Zone groups
76 may have one or more secondary or non-master zones.
77
78 In this guide, the ``rgw1`` host will serve as the master zone of the
79 master zone group; and, the ``rgw2`` host will serve as the secondary zone
80 of the master zone group.
81
82 See `Pools`_ for instructions on creating and tuning pools for Ceph
83 Object Storage.
84
85 See `Sync Policy Config`_ for instructions on defining fine grained bucket sync
86 policy rules.
87
88 .. _master-zone-label:
89
90 Configuring a Master Zone
91 =========================
92
93 All gateways in a multi-site configuration will retrieve their
94 configuration from a ``ceph-radosgw`` daemon on a host within the master
95 zone group and master zone. To configure your gateways in a multi-site
96 configuration, choose a ``ceph-radosgw`` instance to configure the
97 master zone group and master zone.
98
99 Create a Realm
100 --------------
101
102 A realm contains the multi-site configuration of zone groups and zones
103 and also serves to enforce a globally unique namespace within the realm.
104
105 Create a new realm for the multi-site configuration by opening a command
106 line interface on a host identified to serve in the master zone group
107 and zone. Then, execute the following:
108
109 ::
110
111 # radosgw-admin realm create --rgw-realm={realm-name} [--default]
112
113 For example:
114
115 ::
116
117 # radosgw-admin realm create --rgw-realm=movies --default
118
119 If the cluster will have a single realm, specify the ``--default`` flag.
120 If ``--default`` is specified, ``radosgw-admin`` will use this realm by
121 default. If ``--default`` is not specified, adding zone-groups and zones
122 requires specifying either the ``--rgw-realm`` flag or the
123 ``--realm-id`` flag to identify the realm when adding zone groups and
124 zones.
125
126 After creating the realm, ``radosgw-admin`` will echo back the realm
127 configuration. For example:
128
129 ::
130
131 {
132 "id": "0956b174-fe14-4f97-8b50-bb7ec5e1cf62",
133 "name": "movies",
134 "current_period": "1950b710-3e63-4c41-a19e-46a715000980",
135 "epoch": 1
136 }
137
138 .. note:: Ceph generates a unique ID for the realm, which allows the renaming
139 of a realm if the need arises.
140
141 Create a Master Zone Group
142 --------------------------
143
144 A realm must have at least one zone group, which will serve as the
145 master zone group for the realm.
146
147 Create a new master zone group for the multi-site configuration by
148 opening a command line interface on a host identified to serve in the
149 master zone group and zone. Then, execute the following:
150
151 ::
152
153 # radosgw-admin zonegroup create --rgw-zonegroup={name} --endpoints={url} [--rgw-realm={realm-name}|--realm-id={realm-id}] --master --default
154
155 For example:
156
157 ::
158
159 # radosgw-admin zonegroup create --rgw-zonegroup=us --endpoints=http://rgw1:80 --rgw-realm=movies --master --default
160
161 If the realm will only have a single zone group, specify the
162 ``--default`` flag. If ``--default`` is specified, ``radosgw-admin``
163 will use this zone group by default when adding new zones. If
164 ``--default`` is not specified, adding zones will require either the
165 ``--rgw-zonegroup`` flag or the ``--zonegroup-id`` flag to identify the
166 zone group when adding or modifying zones.
167
168 After creating the master zone group, ``radosgw-admin`` will echo back
169 the zone group configuration. For example:
170
171 ::
172
173 {
174 "id": "f1a233f5-c354-4107-b36c-df66126475a6",
175 "name": "us",
176 "api_name": "us",
177 "is_master": "true",
178 "endpoints": [
179 "http:\/\/rgw1:80"
180 ],
181 "hostnames": [],
182 "hostnames_s3webzone": [],
183 "master_zone": "",
184 "zones": [],
185 "placement_targets": [],
186 "default_placement": "",
187 "realm_id": "0956b174-fe14-4f97-8b50-bb7ec5e1cf62"
188 }
189
190 Create a Master Zone
191 --------------------
192
193 .. important:: Zones must be created on a Ceph Object Gateway node that will be
194 within the zone.
195
196 Create a new master zone for the multi-site configuration by opening a
197 command line interface on a host identified to serve in the master zone
198 group and zone. Then, execute the following:
199
200 ::
201
202 # radosgw-admin zone create --rgw-zonegroup={zone-group-name} \
203 --rgw-zone={zone-name} \
204 --master --default \
205 --endpoints={http://fqdn}[,{http://fqdn}]
206
207
208 For example:
209
210 ::
211
212 # radosgw-admin zone create --rgw-zonegroup=us --rgw-zone=us-east \
213 --master --default \
214 --endpoints={http://fqdn}[,{http://fqdn}]
215
216
217 .. note:: The ``--access-key`` and ``--secret`` aren’t specified. These
218 settings will be added to the zone once the user is created in the
219 next section.
220
221 .. important:: The following steps assume a multi-site configuration using newly
222 installed systems that aren’t storing data yet. DO NOT DELETE the
223 ``default`` zone and its pools if you are already using it to store
224 data, or the data will be deleted and unrecoverable.
225
226 Delete Default Zone Group and Zone
227 ----------------------------------
228
229 Delete the ``default`` zone if it exists. Make sure to remove it from
230 the default zone group first.
231
232 ::
233
234 # radosgw-admin zonegroup remove --rgw-zonegroup=default --rgw-zone=default
235 # radosgw-admin period update --commit
236 # radosgw-admin zone rm --rgw-zone=default
237 # radosgw-admin period update --commit
238 # radosgw-admin zonegroup delete --rgw-zonegroup=default
239 # radosgw-admin period update --commit
240
241 Finally, delete the ``default`` pools in your Ceph storage cluster if
242 they exist.
243
244 .. important:: The following step assumes a multi-site configuration using newly
245 installed systems that aren’t currently storing data. DO NOT DELETE
246 the ``default`` zone group if you are already using it to store
247 data.
248
249 ::
250
251 # ceph osd pool rm default.rgw.control default.rgw.control --yes-i-really-really-mean-it
252 # ceph osd pool rm default.rgw.data.root default.rgw.data.root --yes-i-really-really-mean-it
253 # ceph osd pool rm default.rgw.gc default.rgw.gc --yes-i-really-really-mean-it
254 # ceph osd pool rm default.rgw.log default.rgw.log --yes-i-really-really-mean-it
255 # ceph osd pool rm default.rgw.users.uid default.rgw.users.uid --yes-i-really-really-mean-it
256
257 Create a System User
258 --------------------
259
260 The ``ceph-radosgw`` daemons must authenticate before pulling realm and
261 period information. In the master zone, create a system user to
262 facilitate authentication between daemons.
263
264 ::
265
266 # radosgw-admin user create --uid="{user-name}" --display-name="{Display Name}" --system
267
268 For example:
269
270 ::
271
272 # radosgw-admin user create --uid="synchronization-user" --display-name="Synchronization User" --system
273
274 Make a note of the ``access_key`` and ``secret_key``, as the secondary
275 zones will require them to authenticate with the master zone.
276
277 Finally, add the system user to the master zone.
278
279 ::
280
281 # radosgw-admin zone modify --rgw-zone=us-east --access-key={access-key} --secret={secret}
282 # radosgw-admin period update --commit
283
284 Update the Period
285 -----------------
286
287 After updating the master zone configuration, update the period.
288
289 ::
290
291 # radosgw-admin period update --commit
292
293 .. note:: Updating the period changes the epoch, and ensures that other zones
294 will receive the updated configuration.
295
296 Update the Ceph Configuration File
297 ----------------------------------
298
299 Update the Ceph configuration file on master zone hosts by adding the
300 ``rgw_zone`` configuration option and the name of the master zone to the
301 instance entry.
302
303 ::
304
305 [client.rgw.{instance-name}]
306 ...
307 rgw_zone={zone-name}
308
309 For example:
310
311 ::
312
313 [client.rgw.rgw1]
314 host = rgw1
315 rgw frontends = "civetweb port=80"
316 rgw_zone=us-east
317
318 Start the Gateway
319 -----------------
320
321 On the object gateway host, start and enable the Ceph Object Gateway
322 service:
323
324 ::
325
326 # systemctl start ceph-radosgw@rgw.`hostname -s`
327 # systemctl enable ceph-radosgw@rgw.`hostname -s`
328
329 .. _secondary-zone-label:
330
331 Configure Secondary Zones
332 =========================
333
334 Zones within a zone group replicate all data to ensure that each zone
335 has the same data. When creating the secondary zone, execute all of the
336 following operations on a host identified to serve the secondary zone.
337
338 .. note:: To add a third zone, follow the same procedures as for adding the
339 secondary zone. Use different zone name.
340
341 .. important:: You must execute metadata operations, such as user creation, on a
342 host within the master zone. The master zone and the secondary zone
343 can receive bucket operations, but the secondary zone redirects
344 bucket operations to the master zone. If the master zone is down,
345 bucket operations will fail.
346
347 Pull the Realm
348 --------------
349
350 Using the URL path, access key and secret of the master zone in the
351 master zone group, pull the realm configuration to the host. To pull a
352 non-default realm, specify the realm using the ``--rgw-realm`` or
353 ``--realm-id`` configuration options.
354
355 ::
356
357 # radosgw-admin realm pull --url={url-to-master-zone-gateway} --access-key={access-key} --secret={secret}
358
359 .. note:: Pulling the realm also retrieves the remote's current period
360 configuration, and makes it the current period on this host as well.
361
362 If this realm is the default realm or the only realm, make the realm the
363 default realm.
364
365 ::
366
367 # radosgw-admin realm default --rgw-realm={realm-name}
368
369 Create a Secondary Zone
370 -----------------------
371
372 .. important:: Zones must be created on a Ceph Object Gateway node that will be
373 within the zone.
374
375 Create a secondary zone for the multi-site configuration by opening a
376 command line interface on a host identified to serve the secondary zone.
377 Specify the zone group ID, the new zone name and an endpoint for the
378 zone. **DO NOT** use the ``--master`` or ``--default`` flags. In Kraken,
379 all zones run in an active-active configuration by
380 default; that is, a gateway client may write data to any zone and the
381 zone will replicate the data to all other zones within the zone group.
382 If the secondary zone should not accept write operations, specify the
383 ``--read-only`` flag to create an active-passive configuration between
384 the master zone and the secondary zone. Additionally, provide the
385 ``access_key`` and ``secret_key`` of the generated system user stored in
386 the master zone of the master zone group. Execute the following:
387
388 ::
389
390 # radosgw-admin zone create --rgw-zonegroup={zone-group-name}\
391 --rgw-zone={zone-name} --endpoints={url} \
392 --access-key={system-key} --secret={secret}\
393 --endpoints=http://{fqdn}:80 \
394 [--read-only]
395
396 For example:
397
398 ::
399
400 # radosgw-admin zone create --rgw-zonegroup=us --rgw-zone=us-west \
401 --access-key={system-key} --secret={secret} \
402 --endpoints=http://rgw2:80
403
404 .. important:: The following steps assume a multi-site configuration using newly
405 installed systems that aren’t storing data. **DO NOT DELETE** the
406 ``default`` zone and its pools if you are already using it to store
407 data, or the data will be lost and unrecoverable.
408
409 Delete the default zone if needed.
410
411 ::
412
413 # radosgw-admin zone rm --rgw-zone=default
414
415 Finally, delete the default pools in your Ceph storage cluster if
416 needed.
417
418 ::
419
420 # ceph osd pool rm default.rgw.control default.rgw.control --yes-i-really-really-mean-it
421 # ceph osd pool rm default.rgw.data.root default.rgw.data.root --yes-i-really-really-mean-it
422 # ceph osd pool rm default.rgw.gc default.rgw.gc --yes-i-really-really-mean-it
423 # ceph osd pool rm default.rgw.log default.rgw.log --yes-i-really-really-mean-it
424 # ceph osd pool rm default.rgw.users.uid default.rgw.users.uid --yes-i-really-really-mean-it
425
426 Update the Ceph Configuration File
427 ----------------------------------
428
429 Update the Ceph configuration file on the secondary zone hosts by adding
430 the ``rgw_zone`` configuration option and the name of the secondary zone
431 to the instance entry.
432
433 ::
434
435 [client.rgw.{instance-name}]
436 ...
437 rgw_zone={zone-name}
438
439 For example:
440
441 ::
442
443 [client.rgw.rgw2]
444 host = rgw2
445 rgw frontends = "civetweb port=80"
446 rgw_zone=us-west
447
448 Update the Period
449 -----------------
450
451 After updating the master zone configuration, update the period.
452
453 ::
454
455 # radosgw-admin period update --commit
456
457 .. note:: Updating the period changes the epoch, and ensures that other zones
458 will receive the updated configuration.
459
460 Start the Gateway
461 -----------------
462
463 On the object gateway host, start and enable the Ceph Object Gateway
464 service:
465
466 ::
467
468 # systemctl start ceph-radosgw@rgw.`hostname -s`
469 # systemctl enable ceph-radosgw@rgw.`hostname -s`
470
471 Check Synchronization Status
472 ----------------------------
473
474 Once the secondary zone is up and running, check the synchronization
475 status. Synchronization copies users and buckets created in the master
476 zone to the secondary zone.
477
478 ::
479
480 # radosgw-admin sync status
481
482 The output will provide the status of synchronization operations. For
483 example:
484
485 ::
486
487 realm f3239bc5-e1a8-4206-a81d-e1576480804d (earth)
488 zonegroup c50dbb7e-d9ce-47cc-a8bb-97d9b399d388 (us)
489 zone 4c453b70-4a16-4ce8-8185-1893b05d346e (us-west)
490 metadata sync syncing
491 full sync: 0/64 shards
492 metadata is caught up with master
493 incremental sync: 64/64 shards
494 data sync source: 1ee9da3e-114d-4ae3-a8a4-056e8a17f532 (us-east)
495 syncing
496 full sync: 0/128 shards
497 incremental sync: 128/128 shards
498 data is caught up with source
499
500 .. note:: Secondary zones accept bucket operations; however, secondary zones
501 redirect bucket operations to the master zone and then synchronize
502 with the master zone to receive the result of the bucket operations.
503 If the master zone is down, bucket operations executed on the
504 secondary zone will fail, but object operations should succeed.
505
506
507 Maintenance
508 ===========
509
510 Checking the Sync Status
511 ------------------------
512
513 Information about the replication status of a zone can be queried with::
514
515 $ radosgw-admin sync status
516 realm b3bc1c37-9c44-4b89-a03b-04c269bea5da (earth)
517 zonegroup f54f9b22-b4b6-4a0e-9211-fa6ac1693f49 (us)
518 zone adce11c9-b8ed-4a90-8bc5-3fc029ff0816 (us-2)
519 metadata sync syncing
520 full sync: 0/64 shards
521 incremental sync: 64/64 shards
522 metadata is behind on 1 shards
523 oldest incremental change not applied: 2017-03-22 10:20:00.0.881361s
524 data sync source: 341c2d81-4574-4d08-ab0f-5a2a7b168028 (us-1)
525 syncing
526 full sync: 0/128 shards
527 incremental sync: 128/128 shards
528 data is caught up with source
529 source: 3b5d1a3f-3f27-4e4a-8f34-6072d4bb1275 (us-3)
530 syncing
531 full sync: 0/128 shards
532 incremental sync: 128/128 shards
533 data is caught up with source
534
535 Changing the Metadata Master Zone
536 ---------------------------------
537
538 .. important:: Care must be taken when changing which zone is the metadata
539 master. If a zone has not finished syncing metadata from the current master
540 zone, it will be unable to serve any remaining entries when promoted to
541 master and those changes will be lost. For this reason, waiting for a
542 zone's ``radosgw-admin sync status`` to catch up on metadata sync before
543 promoting it to master is recommended.
544
545 Similarly, if changes to metadata are being processed by the current master
546 zone while another zone is being promoted to master, those changes are
547 likely to be lost. To avoid this, shutting down any ``radosgw`` instances
548 on the previous master zone is recommended. After promoting another zone,
549 its new period can be fetched with ``radosgw-admin period pull`` and the
550 gateway(s) can be restarted.
551
552 To promote a zone (for example, zone ``us-2`` in zonegroup ``us``) to metadata
553 master, run the following commands on that zone::
554
555 $ radosgw-admin zone modify --rgw-zone=us-2 --master
556 $ radosgw-admin zonegroup modify --rgw-zonegroup=us --master
557 $ radosgw-admin period update --commit
558
559 This will generate a new period, and the radosgw instance(s) in zone ``us-2``
560 will send this period to other zones.
561
562 Failover and Disaster Recovery
563 ==============================
564
565 If the master zone should fail, failover to the secondary zone for
566 disaster recovery.
567
568 1. Make the secondary zone the master and default zone. For example:
569
570 ::
571
572 # radosgw-admin zone modify --rgw-zone={zone-name} --master --default
573
574 By default, Ceph Object Gateway will run in an active-active
575 configuration. If the cluster was configured to run in an
576 active-passive configuration, the secondary zone is a read-only zone.
577 Remove the ``--read-only`` status to allow the zone to receive write
578 operations. For example:
579
580 ::
581
582 # radosgw-admin zone modify --rgw-zone={zone-name} --master --default \
583 --read-only=false
584
585 2. Update the period to make the changes take effect.
586
587 ::
588
589 # radosgw-admin period update --commit
590
591 3. Finally, restart the Ceph Object Gateway.
592
593 ::
594
595 # systemctl restart ceph-radosgw@rgw.`hostname -s`
596
597 If the former master zone recovers, revert the operation.
598
599 1. From the recovered zone, pull the latest realm configuration
600 from the current master zone.
601
602 ::
603
604 # radosgw-admin realm pull --url={url-to-master-zone-gateway} \
605 --access-key={access-key} --secret={secret}
606
607 2. Make the recovered zone the master and default zone.
608
609 ::
610
611 # radosgw-admin zone modify --rgw-zone={zone-name} --master --default
612
613 3. Update the period to make the changes take effect.
614
615 ::
616
617 # radosgw-admin period update --commit
618
619 4. Then, restart the Ceph Object Gateway in the recovered zone.
620
621 ::
622
623 # systemctl restart ceph-radosgw@rgw.`hostname -s`
624
625 5. If the secondary zone needs to be a read-only configuration, update
626 the secondary zone.
627
628 ::
629
630 # radosgw-admin zone modify --rgw-zone={zone-name} --read-only
631
632 6. Update the period to make the changes take effect.
633
634 ::
635
636 # radosgw-admin period update --commit
637
638 7. Finally, restart the Ceph Object Gateway in the secondary zone.
639
640 ::
641
642 # systemctl restart ceph-radosgw@rgw.`hostname -s`
643
644 .. _rgw-multisite-migrate-from-single-site:
645
646 Migrating a Single Site System to Multi-Site
647 ============================================
648
649 To migrate from a single site system with a ``default`` zone group and
650 zone to a multi site system, use the following steps:
651
652 1. Create a realm. Replace ``<name>`` with the realm name.
653
654 ::
655
656 # radosgw-admin realm create --rgw-realm=<name> --default
657
658 2. Rename the default zone and zonegroup. Replace ``<name>`` with the
659 zonegroup or zone name.
660
661 ::
662
663 # radosgw-admin zonegroup rename --rgw-zonegroup default --zonegroup-new-name=<name>
664 # radosgw-admin zone rename --rgw-zone default --zone-new-name us-east-1 --rgw-zonegroup=<name>
665
666 3. Configure the master zonegroup. Replace ``<name>`` with the realm or
667 zonegroup name. Replace ``<fqdn>`` with the fully qualified domain
668 name(s) in the zonegroup.
669
670 ::
671
672 # radosgw-admin zonegroup modify --rgw-realm=<name> --rgw-zonegroup=<name> --endpoints http://<fqdn>:80 --master --default
673
674 4. Configure the master zone. Replace ``<name>`` with the realm,
675 zonegroup or zone name. Replace ``<fqdn>`` with the fully qualified
676 domain name(s) in the zonegroup.
677
678 ::
679
680 # radosgw-admin zone modify --rgw-realm=<name> --rgw-zonegroup=<name> \
681 --rgw-zone=<name> --endpoints http://<fqdn>:80 \
682 --access-key=<access-key> --secret=<secret-key> \
683 --master --default
684
685 5. Create a system user. Replace ``<user-id>`` with the username.
686 Replace ``<display-name>`` with a display name. It may contain
687 spaces.
688
689 ::
690
691 # radosgw-admin user create --uid=<user-id> --display-name="<display-name>"\
692 --access-key=<access-key> --secret=<secret-key> --system
693
694 6. Commit the updated configuration.
695
696 ::
697
698 # radosgw-admin period update --commit
699
700 7. Finally, restart the Ceph Object Gateway.
701
702 ::
703
704 # systemctl restart ceph-radosgw@rgw.`hostname -s`
705
706 After completing this procedure, proceed to `Configure a Secondary
707 Zone <#configure-secondary-zones>`__ to create a secondary zone
708 in the master zone group.
709
710
711 Multi-Site Configuration Reference
712 ==================================
713
714 The following sections provide additional details and command-line
715 usage for realms, periods, zone groups and zones.
716
717 Realms
718 ------
719
720 A realm represents a globally unique namespace consisting of one or more
721 zonegroups containing one or more zones, and zones containing buckets,
722 which in turn contain objects. A realm enables the Ceph Object Gateway
723 to support multiple namespaces and their configuration on the same
724 hardware.
725
726 A realm contains the notion of periods. Each period represents the state
727 of the zone group and zone configuration in time. Each time you make a
728 change to a zonegroup or zone, update the period and commit it.
729
730 By default, the Ceph Object Gateway does not create a realm
731 for backward compatibility with Infernalis and earlier releases.
732 However, as a best practice, we recommend creating realms for new
733 clusters.
734
735 Create a Realm
736 ~~~~~~~~~~~~~~
737
738 To create a realm, execute ``realm create`` and specify the realm name.
739 If the realm is the default, specify ``--default``.
740
741 ::
742
743 # radosgw-admin realm create --rgw-realm={realm-name} [--default]
744
745 For example:
746
747 ::
748
749 # radosgw-admin realm create --rgw-realm=movies --default
750
751 By specifying ``--default``, the realm will be called implicitly with
752 each ``radosgw-admin`` call unless ``--rgw-realm`` and the realm name
753 are explicitly provided.
754
755 Make a Realm the Default
756 ~~~~~~~~~~~~~~~~~~~~~~~~
757
758 One realm in the list of realms should be the default realm. There may
759 be only one default realm. If there is only one realm and it wasn’t
760 specified as the default realm when it was created, make it the default
761 realm. Alternatively, to change which realm is the default, execute:
762
763 ::
764
765 # radosgw-admin realm default --rgw-realm=movies
766
767 .. note:: When the realm is default, the command line assumes
768 ``--rgw-realm=<realm-name>`` as an argument.
769
770 Delete a Realm
771 ~~~~~~~~~~~~~~
772
773 To delete a realm, execute ``realm delete`` and specify the realm name.
774
775 ::
776
777 # radosgw-admin realm delete --rgw-realm={realm-name}
778
779 For example:
780
781 ::
782
783 # radosgw-admin realm delete --rgw-realm=movies
784
785 Get a Realm
786 ~~~~~~~~~~~
787
788 To get a realm, execute ``realm get`` and specify the realm name.
789
790 ::
791
792 #radosgw-admin realm get --rgw-realm=<name>
793
794 For example:
795
796 ::
797
798 # radosgw-admin realm get --rgw-realm=movies [> filename.json]
799
800 The CLI will echo a JSON object with the realm properties.
801
802 ::
803
804 {
805 "id": "0a68d52e-a19c-4e8e-b012-a8f831cb3ebc",
806 "name": "movies",
807 "current_period": "b0c5bbef-4337-4edd-8184-5aeab2ec413b",
808 "epoch": 1
809 }
810
811 Use ``>`` and an output file name to output the JSON object to a file.
812
813 Set a Realm
814 ~~~~~~~~~~~
815
816 To set a realm, execute ``realm set``, specify the realm name, and
817 ``--infile=`` with an input file name.
818
819 ::
820
821 #radosgw-admin realm set --rgw-realm=<name> --infile=<infilename>
822
823 For example:
824
825 ::
826
827 # radosgw-admin realm set --rgw-realm=movies --infile=filename.json
828
829 List Realms
830 ~~~~~~~~~~~
831
832 To list realms, execute ``realm list``.
833
834 ::
835
836 # radosgw-admin realm list
837
838 List Realm Periods
839 ~~~~~~~~~~~~~~~~~~
840
841 To list realm periods, execute ``realm list-periods``.
842
843 ::
844
845 # radosgw-admin realm list-periods
846
847 Pull a Realm
848 ~~~~~~~~~~~~
849
850 To pull a realm from the node containing the master zone group and
851 master zone to a node containing a secondary zone group or zone, execute
852 ``realm pull`` on the node that will receive the realm configuration.
853
854 ::
855
856 # radosgw-admin realm pull --url={url-to-master-zone-gateway} --access-key={access-key} --secret={secret}
857
858 Rename a Realm
859 ~~~~~~~~~~~~~~
860
861 A realm is not part of the period. Consequently, renaming the realm is
862 only applied locally, and will not get pulled with ``realm pull``. When
863 renaming a realm with multiple zones, run the command on each zone. To
864 rename a realm, execute the following:
865
866 ::
867
868 # radosgw-admin realm rename --rgw-realm=<current-name> --realm-new-name=<new-realm-name>
869
870 .. note:: DO NOT use ``realm set`` to change the ``name`` parameter. That
871 changes the internal name only. Specifying ``--rgw-realm`` would
872 still use the old realm name.
873
874 Zone Groups
875 -----------
876
877 The Ceph Object Gateway supports multi-site deployments and a global
878 namespace by using the notion of zone groups. Formerly called a region
879 in Infernalis, a zone group defines the geographic location of one or more Ceph
880 Object Gateway instances within one or more zones.
881
882 Configuring zone groups differs from typical configuration procedures,
883 because not all of the settings end up in a Ceph configuration file. You
884 can list zone groups, get a zone group configuration, and set a zone
885 group configuration.
886
887 Create a Zone Group
888 ~~~~~~~~~~~~~~~~~~~
889
890 Creating a zone group consists of specifying the zone group name.
891 Creating a zone assumes it will live in the default realm unless
892 ``--rgw-realm=<realm-name>`` is specified. If the zonegroup is the
893 default zonegroup, specify the ``--default`` flag. If the zonegroup is
894 the master zonegroup, specify the ``--master`` flag. For example:
895
896 ::
897
898 # radosgw-admin zonegroup create --rgw-zonegroup=<name> [--rgw-realm=<name>][--master] [--default]
899
900
901 .. note:: Use ``zonegroup modify --rgw-zonegroup=<zonegroup-name>`` to modify
902 an existing zone group’s settings.
903
904 Make a Zone Group the Default
905 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
906
907 One zonegroup in the list of zonegroups should be the default zonegroup.
908 There may be only one default zonegroup. If there is only one zonegroup
909 and it wasn’t specified as the default zonegroup when it was created,
910 make it the default zonegroup. Alternatively, to change which zonegroup
911 is the default, execute:
912
913 ::
914
915 # radosgw-admin zonegroup default --rgw-zonegroup=comedy
916
917 .. note:: When the zonegroup is default, the command line assumes
918 ``--rgw-zonegroup=<zonegroup-name>`` as an argument.
919
920 Then, update the period:
921
922 ::
923
924 # radosgw-admin period update --commit
925
926 Add a Zone to a Zone Group
927 ~~~~~~~~~~~~~~~~~~~~~~~~~~
928
929 To add a zone to a zonegroup, execute the following:
930
931 ::
932
933 # radosgw-admin zonegroup add --rgw-zonegroup=<name> --rgw-zone=<name>
934
935 Then, update the period:
936
937 ::
938
939 # radosgw-admin period update --commit
940
941 Remove a Zone from a Zone Group
942 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
943
944 To remove a zone from a zonegroup, execute the following:
945
946 ::
947
948 # radosgw-admin zonegroup remove --rgw-zonegroup=<name> --rgw-zone=<name>
949
950 Then, update the period:
951
952 ::
953
954 # radosgw-admin period update --commit
955
956 Rename a Zone Group
957 ~~~~~~~~~~~~~~~~~~~
958
959 To rename a zonegroup, execute the following:
960
961 ::
962
963 # radosgw-admin zonegroup rename --rgw-zonegroup=<name> --zonegroup-new-name=<name>
964
965 Then, update the period:
966
967 ::
968
969 # radosgw-admin period update --commit
970
971 Delete a Zone Group
972 ~~~~~~~~~~~~~~~~~~~
973
974 To delete a zonegroup, execute the following:
975
976 ::
977
978 # radosgw-admin zonegroup delete --rgw-zonegroup=<name>
979
980 Then, update the period:
981
982 ::
983
984 # radosgw-admin period update --commit
985
986 List Zone Groups
987 ~~~~~~~~~~~~~~~~
988
989 A Ceph cluster contains a list of zone groups. To list the zone groups,
990 execute:
991
992 ::
993
994 # radosgw-admin zonegroup list
995
996 The ``radosgw-admin`` returns a JSON formatted list of zone groups.
997
998 ::
999
1000 {
1001 "default_info": "90b28698-e7c3-462c-a42d-4aa780d24eda",
1002 "zonegroups": [
1003 "us"
1004 ]
1005 }
1006
1007 Get a Zone Group Map
1008 ~~~~~~~~~~~~~~~~~~~~
1009
1010 To list the details of each zone group, execute:
1011
1012 ::
1013
1014 # radosgw-admin zonegroup-map get
1015
1016 .. note:: If you receive a ``failed to read zonegroup map`` error, run
1017 ``radosgw-admin zonegroup-map update`` as ``root`` first.
1018
1019 Get a Zone Group
1020 ~~~~~~~~~~~~~~~~
1021
1022 To view the configuration of a zone group, execute:
1023
1024 ::
1025
1026 radosgw-admin zonegroup get [--rgw-zonegroup=<zonegroup>]
1027
1028 The zone group configuration looks like this:
1029
1030 ::
1031
1032 {
1033 "id": "90b28698-e7c3-462c-a42d-4aa780d24eda",
1034 "name": "us",
1035 "api_name": "us",
1036 "is_master": "true",
1037 "endpoints": [
1038 "http:\/\/rgw1:80"
1039 ],
1040 "hostnames": [],
1041 "hostnames_s3website": [],
1042 "master_zone": "9248cab2-afe7-43d8-a661-a40bf316665e",
1043 "zones": [
1044 {
1045 "id": "9248cab2-afe7-43d8-a661-a40bf316665e",
1046 "name": "us-east",
1047 "endpoints": [
1048 "http:\/\/rgw1"
1049 ],
1050 "log_meta": "true",
1051 "log_data": "true",
1052 "bucket_index_max_shards": 0,
1053 "read_only": "false"
1054 },
1055 {
1056 "id": "d1024e59-7d28-49d1-8222-af101965a939",
1057 "name": "us-west",
1058 "endpoints": [
1059 "http:\/\/rgw2:80"
1060 ],
1061 "log_meta": "false",
1062 "log_data": "true",
1063 "bucket_index_max_shards": 0,
1064 "read_only": "false"
1065 }
1066 ],
1067 "placement_targets": [
1068 {
1069 "name": "default-placement",
1070 "tags": []
1071 }
1072 ],
1073 "default_placement": "default-placement",
1074 "realm_id": "ae031368-8715-4e27-9a99-0c9468852cfe"
1075 }
1076
1077 Set a Zone Group
1078 ~~~~~~~~~~~~~~~~
1079
1080 Defining a zone group consists of creating a JSON object, specifying at
1081 least the required settings:
1082
1083 1. ``name``: The name of the zone group. Required.
1084
1085 2. ``api_name``: The API name for the zone group. Optional.
1086
1087 3. ``is_master``: Determines if the zone group is the master zone group.
1088 Required. **note:** You can only have one master zone group.
1089
1090 4. ``endpoints``: A list of all the endpoints in the zone group. For
1091 example, you may use multiple domain names to refer to the same zone
1092 group. Remember to escape the forward slashes (``\/``). You may also
1093 specify a port (``fqdn:port``) for each endpoint. Optional.
1094
1095 5. ``hostnames``: A list of all the hostnames in the zone group. For
1096 example, you may use multiple domain names to refer to the same zone
1097 group. Optional. The ``rgw dns name`` setting will automatically be
1098 included in this list. You should restart the gateway daemon(s) after
1099 changing this setting.
1100
1101 6. ``master_zone``: The master zone for the zone group. Optional. Uses
1102 the default zone if not specified. **note:** You can only have one
1103 master zone per zone group.
1104
1105 7. ``zones``: A list of all zones within the zone group. Each zone has a
1106 name (required), a list of endpoints (optional), and whether or not
1107 the gateway will log metadata and data operations (false by default).
1108
1109 8. ``placement_targets``: A list of placement targets (optional). Each
1110 placement target contains a name (required) for the placement target
1111 and a list of tags (optional) so that only users with the tag can use
1112 the placement target (i.e., the user’s ``placement_tags`` field in
1113 the user info).
1114
1115 9. ``default_placement``: The default placement target for the object
1116 index and object data. Set to ``default-placement`` by default. You
1117 may also set a per-user default placement in the user info for each
1118 user.
1119
1120 To set a zone group, create a JSON object consisting of the required
1121 fields, save the object to a file (e.g., ``zonegroup.json``); then,
1122 execute the following command:
1123
1124 ::
1125
1126 # radosgw-admin zonegroup set --infile zonegroup.json
1127
1128 Where ``zonegroup.json`` is the JSON file you created.
1129
1130 .. important:: The ``default`` zone group ``is_master`` setting is ``true`` by
1131 default. If you create a new zone group and want to make it the
1132 master zone group, you must either set the ``default`` zone group
1133 ``is_master`` setting to ``false``, or delete the ``default`` zone
1134 group.
1135
1136 Finally, update the period:
1137
1138 ::
1139
1140 # radosgw-admin period update --commit
1141
1142 Set a Zone Group Map
1143 ~~~~~~~~~~~~~~~~~~~~
1144
1145 Setting a zone group map consists of creating a JSON object consisting
1146 of one or more zone groups, and setting the ``master_zonegroup`` for the
1147 cluster. Each zone group in the zone group map consists of a key/value
1148 pair, where the ``key`` setting is equivalent to the ``name`` setting
1149 for an individual zone group configuration, and the ``val`` is a JSON
1150 object consisting of an individual zone group configuration.
1151
1152 You may only have one zone group with ``is_master`` equal to ``true``,
1153 and it must be specified as the ``master_zonegroup`` at the end of the
1154 zone group map. The following JSON object is an example of a default
1155 zone group map.
1156
1157 ::
1158
1159 {
1160 "zonegroups": [
1161 {
1162 "key": "90b28698-e7c3-462c-a42d-4aa780d24eda",
1163 "val": {
1164 "id": "90b28698-e7c3-462c-a42d-4aa780d24eda",
1165 "name": "us",
1166 "api_name": "us",
1167 "is_master": "true",
1168 "endpoints": [
1169 "http:\/\/rgw1:80"
1170 ],
1171 "hostnames": [],
1172 "hostnames_s3website": [],
1173 "master_zone": "9248cab2-afe7-43d8-a661-a40bf316665e",
1174 "zones": [
1175 {
1176 "id": "9248cab2-afe7-43d8-a661-a40bf316665e",
1177 "name": "us-east",
1178 "endpoints": [
1179 "http:\/\/rgw1"
1180 ],
1181 "log_meta": "true",
1182 "log_data": "true",
1183 "bucket_index_max_shards": 0,
1184 "read_only": "false"
1185 },
1186 {
1187 "id": "d1024e59-7d28-49d1-8222-af101965a939",
1188 "name": "us-west",
1189 "endpoints": [
1190 "http:\/\/rgw2:80"
1191 ],
1192 "log_meta": "false",
1193 "log_data": "true",
1194 "bucket_index_max_shards": 0,
1195 "read_only": "false"
1196 }
1197 ],
1198 "placement_targets": [
1199 {
1200 "name": "default-placement",
1201 "tags": []
1202 }
1203 ],
1204 "default_placement": "default-placement",
1205 "realm_id": "ae031368-8715-4e27-9a99-0c9468852cfe"
1206 }
1207 }
1208 ],
1209 "master_zonegroup": "90b28698-e7c3-462c-a42d-4aa780d24eda",
1210 "bucket_quota": {
1211 "enabled": false,
1212 "max_size_kb": -1,
1213 "max_objects": -1
1214 },
1215 "user_quota": {
1216 "enabled": false,
1217 "max_size_kb": -1,
1218 "max_objects": -1
1219 }
1220 }
1221
1222 To set a zone group map, execute the following:
1223
1224 ::
1225
1226 # radosgw-admin zonegroup-map set --infile zonegroupmap.json
1227
1228 Where ``zonegroupmap.json`` is the JSON file you created. Ensure that
1229 you have zones created for the ones specified in the zone group map.
1230 Finally, update the period.
1231
1232 ::
1233
1234 # radosgw-admin period update --commit
1235
1236 Zones
1237 -----
1238
1239 Ceph Object Gateway supports the notion of zones. A zone defines a
1240 logical group consisting of one or more Ceph Object Gateway instances.
1241
1242 Configuring zones differs from typical configuration procedures, because
1243 not all of the settings end up in a Ceph configuration file. You can
1244 list zones, get a zone configuration and set a zone configuration.
1245
1246 Create a Zone
1247 ~~~~~~~~~~~~~
1248
1249 To create a zone, specify a zone name. If it is a master zone, specify
1250 the ``--master`` option. Only one zone in a zone group may be a master
1251 zone. To add the zone to a zonegroup, specify the ``--rgw-zonegroup``
1252 option with the zonegroup name.
1253
1254 ::
1255
1256 # radosgw-admin zone create --rgw-zone=<name> \
1257 [--zonegroup=<zonegroup-name]\
1258 [--endpoints=<endpoint>[,<endpoint>] \
1259 [--master] [--default] \
1260 --access-key $SYSTEM_ACCESS_KEY --secret $SYSTEM_SECRET_KEY
1261
1262 Then, update the period:
1263
1264 ::
1265
1266 # radosgw-admin period update --commit
1267
1268 Delete a Zone
1269 ~~~~~~~~~~~~~
1270
1271 To delete zone, first remove it from the zonegroup.
1272
1273 ::
1274
1275 # radosgw-admin zonegroup remove --zonegroup=<name>\
1276 --zone=<name>
1277
1278 Then, update the period:
1279
1280 ::
1281
1282 # radosgw-admin period update --commit
1283
1284 Next, delete the zone. Execute the following:
1285
1286 ::
1287
1288 # radosgw-admin zone rm --rgw-zone<name>
1289
1290 Finally, update the period:
1291
1292 ::
1293
1294 # radosgw-admin period update --commit
1295
1296 .. important:: Do not delete a zone without removing it from a zone group first.
1297 Otherwise, updating the period will fail.
1298
1299 If the pools for the deleted zone will not be used anywhere else,
1300 consider deleting the pools. Replace ``<del-zone>`` in the example below
1301 with the deleted zone’s name.
1302
1303 .. important:: Only delete the pools with prepended zone names. Deleting the root
1304 pool, such as, ``.rgw.root`` will remove all of the system’s
1305 configuration.
1306
1307 .. important:: Once the pools are deleted, all of the data within them are deleted
1308 in an unrecoverable manner. Only delete the pools if the pool
1309 contents are no longer needed.
1310
1311 ::
1312
1313 # ceph osd pool rm <del-zone>.rgw.control <del-zone>.rgw.control --yes-i-really-really-mean-it
1314 # ceph osd pool rm <del-zone>.rgw.data.root <del-zone>.rgw.data.root --yes-i-really-really-mean-it
1315 # ceph osd pool rm <del-zone>.rgw.gc <del-zone>.rgw.gc --yes-i-really-really-mean-it
1316 # ceph osd pool rm <del-zone>.rgw.log <del-zone>.rgw.log --yes-i-really-really-mean-it
1317 # ceph osd pool rm <del-zone>.rgw.users.uid <del-zone>.rgw.users.uid --yes-i-really-really-mean-it
1318
1319 Modify a Zone
1320 ~~~~~~~~~~~~~
1321
1322 To modify a zone, specify the zone name and the parameters you wish to
1323 modify.
1324
1325 ::
1326
1327 # radosgw-admin zone modify [options]
1328
1329 Where ``[options]``:
1330
1331 - ``--access-key=<key>``
1332 - ``--secret/--secret-key=<key>``
1333 - ``--master``
1334 - ``--default``
1335 - ``--endpoints=<list>``
1336
1337 Then, update the period:
1338
1339 ::
1340
1341 # radosgw-admin period update --commit
1342
1343 List Zones
1344 ~~~~~~~~~~
1345
1346 As ``root``, to list the zones in a cluster, execute:
1347
1348 ::
1349
1350 # radosgw-admin zone list
1351
1352 Get a Zone
1353 ~~~~~~~~~~
1354
1355 As ``root``, to get the configuration of a zone, execute:
1356
1357 ::
1358
1359 # radosgw-admin zone get [--rgw-zone=<zone>]
1360
1361 The ``default`` zone looks like this:
1362
1363 ::
1364
1365 { "domain_root": ".rgw",
1366 "control_pool": ".rgw.control",
1367 "gc_pool": ".rgw.gc",
1368 "log_pool": ".log",
1369 "intent_log_pool": ".intent-log",
1370 "usage_log_pool": ".usage",
1371 "user_keys_pool": ".users",
1372 "user_email_pool": ".users.email",
1373 "user_swift_pool": ".users.swift",
1374 "user_uid_pool": ".users.uid",
1375 "system_key": { "access_key": "", "secret_key": ""},
1376 "placement_pools": [
1377 { "key": "default-placement",
1378 "val": { "index_pool": ".rgw.buckets.index",
1379 "data_pool": ".rgw.buckets"}
1380 }
1381 ]
1382 }
1383
1384 Set a Zone
1385 ~~~~~~~~~~
1386
1387 Configuring a zone involves specifying a series of Ceph Object Gateway
1388 pools. For consistency, we recommend using a pool prefix that is the
1389 same as the zone name. See
1390 `Pools <http://docs.ceph.com/docs/master/rados/operations/pools/#pools>`__
1391 for details of configuring pools.
1392
1393 To set a zone, create a JSON object consisting of the pools, save the
1394 object to a file (e.g., ``zone.json``); then, execute the following
1395 command, replacing ``{zone-name}`` with the name of the zone:
1396
1397 ::
1398
1399 # radosgw-admin zone set --rgw-zone={zone-name} --infile zone.json
1400
1401 Where ``zone.json`` is the JSON file you created.
1402
1403 Then, as ``root``, update the period:
1404
1405 ::
1406
1407 # radosgw-admin period update --commit
1408
1409 Rename a Zone
1410 ~~~~~~~~~~~~~
1411
1412 To rename a zone, specify the zone name and the new zone name.
1413
1414 ::
1415
1416 # radosgw-admin zone rename --rgw-zone=<name> --zone-new-name=<name>
1417
1418 Then, update the period:
1419
1420 ::
1421
1422 # radosgw-admin period update --commit
1423
1424 Zone Group and Zone Settings
1425 ----------------------------
1426
1427 When configuring a default zone group and zone, the pool name includes
1428 the zone name. For example:
1429
1430 - ``default.rgw.control``
1431
1432 To change the defaults, include the following settings in your Ceph
1433 configuration file under each ``[client.radosgw.{instance-name}]``
1434 instance.
1435
1436 +-------------------------------------+-----------------------------------+---------+-----------------------+
1437 | Name | Description | Type | Default |
1438 +=====================================+===================================+=========+=======================+
1439 | ``rgw_zone`` | The name of the zone for the | String | None |
1440 | | gateway instance. | | |
1441 +-------------------------------------+-----------------------------------+---------+-----------------------+
1442 | ``rgw_zonegroup`` | The name of the zone group for | String | None |
1443 | | the gateway instance. | | |
1444 +-------------------------------------+-----------------------------------+---------+-----------------------+
1445 | ``rgw_zonegroup_root_pool`` | The root pool for the zone group. | String | ``.rgw.root`` |
1446 +-------------------------------------+-----------------------------------+---------+-----------------------+
1447 | ``rgw_zone_root_pool`` | The root pool for the zone. | String | ``.rgw.root`` |
1448 +-------------------------------------+-----------------------------------+---------+-----------------------+
1449 | ``rgw_default_zone_group_info_oid`` | The OID for storing the default | String | ``default.zonegroup`` |
1450 | | zone group. We do not recommend | | |
1451 | | changing this setting. | | |
1452 +-------------------------------------+-----------------------------------+---------+-----------------------+
1453
1454
1455 .. _`Pools`: ../pools
1456 .. _`Sync Policy Config`: ../multisite-sync-policy