summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorFrode Nordahl <frode.nordahl@canonical.com>2018-06-05 13:52:05 +0200
committerFrode Nordahl <frode.nordahl@canonical.com>2018-06-05 13:52:05 +0200
commit5c0bef38d3051d4244f50c756998f3d258b4b57b (patch)
tree305c3456defccbf51c987b7325634a437febceed
parenta8b50bff6f23411b7fcb27c4c8c5a306853419c5 (diff)
Make action descriptions terse, move to README.md
Output of `juju list-action` is at time of this writing formatted in such a way that we should keep description as terse as possible and refer to documentation elsewhere. Change-Id: Ib8e7a4804e696199803b9ac386da7bf02aafd465
Notes
Notes (review): Verified+1: Canonical CI <uosci-testing-bot@ubuntu.com> Code-Review+2: Chris MacNaughton <chris.macnaughton@canonical.com> Code-Review+2: James Page <james.page@canonical.com> Workflow+1: James Page <james.page@canonical.com> Verified+2: Zuul Submitted-by: Zuul Submitted-at: Tue, 05 Jun 2018 15:47:16 +0000 Reviewed-on: https://review.openstack.org/572354 Project: openstack/charm-ceph-osd Branch: refs/heads/master
-rw-r--r--README.md113
-rw-r--r--actions.yaml83
2 files changed, 142 insertions, 54 deletions
diff --git a/README.md b/README.md
index a3f22bf..84c84e9 100644
--- a/README.md
+++ b/README.md
@@ -122,6 +122,119 @@ options::
122new block devices added to the ceph-osd application; existing OSD devices will 122new block devices added to the ceph-osd application; existing OSD devices will
123not be encrypted. 123not be encrypted.
124 124
125Actions
126=======
127The charm offers [actions](https://docs.jujucharms.com/devel/en/actions) which
128may be used to perform operational tasks on individual units.
129
130pause
131-----
132**USE WITH CAUTION** - Set the local osd units in the charm to 'out' but
133does not stop the osds. Unless the osd cluster is set to noout (see below),
134this removes them from the ceph cluster and forces ceph to migrate the PGs
135to other OSDs in the cluster.
136
137From [upstream documentation](http://docs.ceph.com/docs/master/rados/operations/add-or-rm-osds/#removing-the-osd)
138"Do not let your cluster reach its full ratio when removing an OSD.
139 Removing OSDs could cause the cluster to reach or exceed its full ratio."
140
141Also note that for small clusters you may encounter the corner case where
142some PGs remain stuck in the active+remapped state. Refer to the above link
143on how to resolve this.
144
145`pause-health` (on a ceph-mon) unit can be used before pausing a ceph-osd
146unit to stop the cluster rebalancing the data off this ceph-osd unit.
147`pause-health` sets 'noout' on the cluster such that it will not try to
148rebalance the data accross the remaining units.
149
150It is up to the user of the charm to determine whether pause-health should
151be used as it depends on whether the osd is being paused for maintenance or
152to remove it from the cluster completely.
153
154**NOTE** the `pause` action does NOT stop the ceph-osd processes.
155
156resume
157------
158Set the local osd units in the charm to 'in'.
159
160
161list-disks
162----------
163List disks
164
165The 'disks' key is populated with block devices that are known by udev,
166are not mounted and not mentioned in 'osd-journal' configuration option.
167
168The 'blacklist' key is populated with osd-devices in the blacklist stored
169in the local kv store of this specific unit.
170
171The 'non-pristine' key is populated with block devices that are known by
172udev, are not mounted, not mentioned in 'osd-journal' configuration option
173and are currently not eligible for use because of presence of foreign data.
174
175add-disk
176--------
177Add disk(s) to Ceph
178
179#### Parameters
180- `osd-devices` (required)
181 - The devices to format and set up as osd volumes.
182- `bucket`
183 - The name of the bucket in Ceph to add these devices into
184
185blacklist-add-disk
186------------------
187Add disk(s) to blacklist. Blacklisted disks will not be
188initialized for use with Ceph even if listed in the application
189level osd-devices configuration option.
190
191The current blacklist can be viewed with list-disks action.
192
193**NOTE** This action and blacklist will not have any effect on
194already initialized disks.
195
196#### Parameters
197- `osd-devices` (required)
198 - A space-separated list of devices to add to blacklist.
199
200 Each element should be a absolute path to a device node or filesystem
201 directory (the latter is supported for ceph >= 0.56.6).
202
203 Example: '/dev/vdb /var/tmp/test-osd'
204
205blacklist-remove-disk
206---------------------
207Remove disk(s) from blacklist.
208
209#### Parameters
210- `osd-devices` (required)
211 - A space-separated list of devices to remove from blacklist.
212
213 Each element should be a existing entry in the units blacklist.
214 Use list-disks action to list current blacklist entries.
215
216 Example: '/dev/vdb /var/tmp/test-osd'
217
218zap-disk
219--------
220Purge disk of all data and signatures for use by Ceph
221
222This action can be necessary in cases where a Ceph cluster is being
223redeployed as the charm defaults to skipping disks that look like Ceph
224devices in order to preserve data. In order to forcibly redeploy, the
225admin is required to perform this action for each disk to be re-consumed.
226
227In addition to triggering this action, it is required to pass an additional
228parameter option of `i-really-mean-it` to ensure that the
229administrator is aware that this *will* cause data loss on the specified
230device(s)
231
232#### Parameters
233- `devices` (required)
234 - A space-separated list of devices to remove the partition table from.
235- `i-really-mean-it` (required)
236 - This must be toggled to enable actually performing this action
237
125Contact Information 238Contact Information
126=================== 239===================
127 240
diff --git a/actions.yaml b/actions.yaml
index 396dc4c..3be669c 100644
--- a/actions.yaml
+++ b/actions.yaml
@@ -1,44 +1,29 @@
1# NOTE(fnordahl): Output of `juju list-action` is at time of this writing
2# formatted in such a way that we should keep description
3# as terse as possible and refer to documentation elsewhere.
4#
5# Verify with `juju list-action` before proposing/committing
6# changes.
1pause: 7pause:
2 description: | 8 description: |
3 CAUTION - Set the local osd units in the charm to 'out' but does not stop 9 \
4 the osds. Unless the osd cluster is set to noout (see below), this removes 10 USE WITH CAUTION - Mark unit OSDs as 'out'.
5 them from the ceph cluster and forces ceph to migrate the PGs to other OSDs 11 Documentation: https://jujucharms.com/ceph-osd/
6 in the cluster. See the following.
7
8 http://docs.ceph.com/docs/master/rados/operations/add-or-rm-osds/#removing-the-osd
9 "Do not let your cluster reach its full ratio when removing an OSD.
10 Removing OSDs could cause the cluster to reach or exceed its full ratio."
11 Also note that for small clusters you may encounter the corner case where
12 some PGs remain stuck in the active+remapped state. Refer to the above link
13 on how to resolve this.
14
15 pause-health (on a ceph-mon) unit can be used before pausing a ceph-osd
16 unit to stop the cluster rebalancing the data off this ceph-osd unit.
17 pause-health sets 'noout' on the cluster such that it will not try to
18 rebalance the data accross the remaining units.
19
20 It is up to the user of the charm to determine whether pause-health should
21 be used as it depends on whether the osd is being paused for maintenance or
22 to remove it from the cluster completely.
23resume: 12resume:
24 description: | 13 description: |
25 Set the local osd units in the charm to 'in'. Note that the pause option 14 \
26 does NOT stop the osd processes. 15 Set the local osd units in the charm to 'in'.
16 Documentation: https://jujucharms.com/ceph-osd/
27list-disks: 17list-disks:
28 description: | 18 description: |
29 List disks 19 \
30 . 20 List disks.
31 The 'disks' key is populated with block devices that are known by udev, 21 Documentation: https://jujucharms.com/ceph-osd/
32 are not mounted and not mentioned in 'osd-journal' configuration option.
33 .
34 The 'blacklist' key is populated with osd-devices in the blacklist stored
35 in the local kv store of this specific unit.
36 .
37 The 'non-pristine' key is populated with block devices that are known by
38 udev, are not mounted, not mentioned in 'osd-journal' configuration option
39 and are currently not eligible for use because of presence of foreign data.
40add-disk: 22add-disk:
41 description: Add disk(s) to Ceph 23 description: |
24 \
25 Add disk(s) to Ceph.
26 Documentation: https://jujucharms.com/ceph-osd/
42 params: 27 params:
43 osd-devices: 28 osd-devices:
44 type: string 29 type: string
@@ -50,14 +35,9 @@ add-disk:
50 - osd-devices 35 - osd-devices
51blacklist-add-disk: 36blacklist-add-disk:
52 description: | 37 description: |
53 Add disk(s) to blacklist. Blacklisted disks will not be 38 \
54 initialized for use with Ceph even if listed in the application 39 Add disk(s) to blacklist.
55 level osd-devices configuration option. 40 Documentation: https://jujucharms.com/ceph-osd/
56 .
57 The current blacklist can be viewed with list-disks action.
58 .
59 NOTE: This action and blacklist will not have any effect on
60 already initialized disks.
61 params: 41 params:
62 osd-devices: 42 osd-devices:
63 type: string 43 type: string
@@ -71,7 +51,10 @@ blacklist-add-disk:
71 required: 51 required:
72 - osd-devices 52 - osd-devices
73blacklist-remove-disk: 53blacklist-remove-disk:
74 description: Remove disk(s) from blacklist. 54 description: |
55 \
56 Remove disk(s) from blacklist.
57 Documentation: https://jujucharms.com/ceph-osd/
75 params: 58 params:
76 osd-devices: 59 osd-devices:
77 type: string 60 type: string
@@ -86,17 +69,9 @@ blacklist-remove-disk:
86 - osd-devices 69 - osd-devices
87zap-disk: 70zap-disk:
88 description: | 71 description: |
89 Purge disk of all data and signatures for use by Ceph 72 \
90 . 73 USE WITH CAUTION - Purge disk of all data and signatures for use by Ceph.
91 This action can be necessary in cases where a Ceph cluster is being 74 Documentation: https://jujucharms.com/ceph-osd/
92 redeployed as the charm defaults to skipping disks that look like Ceph
93 devices in order to preserve data. In order to forcibly redeploy, the
94 admin is required to perform this action for each disk to be re-consumed.
95 .
96 In addition to triggering this action, it is required to pass an additional
97 parameter option of `i-really-mean-it` to ensure that the
98 administrator is aware that this *will* cause data loss on the specified
99 device(s)
100 params: 75 params:
101 devices: 76 devices:
102 type: string 77 type: string