Commit Graph
5036 Commits
Author SHA1 Message Date
Alexander Trost 6dc508b704 Merge pull request #2318 from galexrt/minio_arbitary_pvcs
Cockroachdb + Minio: Use PersistentVolumeClaimTemplates storage
2018-12-08 00:20:07 +01:00
travisn 304d6445f2 rgw: update the deployment or daemonset when the pod spec changes
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-07 15:11:55 -07:00
Jared Watts 99379308ba Merge pull request #2325 from yanniszark/cassandra-operator-scale-down-implementation
[cassandra operator] Scale Down Implementation
2018-12-07 12:28:06 -08:00
travisn fe4a645844 osd: detect existing osds not to be passed to ceph-volume
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-07 10:50:37 -07:00
travisn 1389cc56db ceph: enable multiple osds per device
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-07 09:29:15 -07:00
Yannis Zarkadas eca28a9469 cassandra: scale down unit tests
Unit tests for the scale down action.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-07 13:46:46 +02:00
Yannis Zarkadas a9178ccb59 cassandra: sidecar-side implementation of scale down
The sidecar now has a control loop, which watches its ClusterIP
Service for the `cassandra.rook.io/decommissioned` label.
If it finds it, then it decommissions and updates the label to
value "true".

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-07 13:46:46 +02:00
Yannis Zarkadas f5a41e9a7c cassandra: controller-side implementation of scale down
Code handling the scale down on the side of the controller.
Also, some new util functions.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-07 13:46:45 +02:00
Yannis Zarkadas 624ff18211 cassandra: add RackConditions to RackStatus
RackConditions are observations about a Cassandra Rack.
Currently, the only condition is MemberLeaving.
Conditions are an api convention of kubernetes.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-07 13:46:45 +02:00
Yannis Zarkadas 9d69894386 cassandra: add label - intent for decommission
We add a new label to our constants, called
`cassandra.rook.io/decommissioned`. When it is present on a
member's ClusterIP Service with the value "false", it expresses
the intent of the controller to decommission the member. When it is
present with the value "true", it means that decommission has completed.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-07 13:46:45 +02:00
Alexander Trost 5fc8584422 Cockroachdb + Minio: Use PersistentVolumeClaimTemplates storage
selection field.

Signed-off-by: Alexander Trost <galexrt@googlemail.com>
2018-12-07 12:20:49 +01:00
travisn e6572e41e6 enable ceph-volume to provision and start filestore osds
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-06 21:28:58 -07:00
travisn 45c53dd5b2 detect if ceph-volume is supported and allow legacy osds to be started
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-06 21:28:58 -07:00
travisn 78b8011226 osds: prototype provisioning with ceph-volume
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-06 21:28:58 -07:00
travisn f7c266f7ec skip sgdisk detection if not found in nautilus image
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-06 21:28:57 -07:00
Travis Nielsen 2d883e5489 Merge pull request #2339 from SUSE/ceph-mon-update-band-aid
Ceph mon: temp change to support #2331 in v0.9
2018-12-06 19:59:08 -07:00
Blaine Gardner 6f68b2d536 Ceph mon: temp change to support #2331 in v0.9
In order to have Ceph updates working for Rook release v0.9, apply a
simple, temporary solution to update the mons without extraneous mon
updates.

Signed-off-by: Blaine Gardner <blaine.gardner@suse.com>
2018-12-06 16:37:42 -07:00
Dmitry Yusupov 18634ccb87 Edgefs codegen and clientset
Signed-off-by: Dmitry Yusupov <dmitry.yusupov@nexenta.com>
2018-12-05 23:05:43 -08:00
Dmitry Yusupov a29477f122 Edgefs operator and CRDs
Signed-off-by: Dmitry Yusupov <dmitry.yusupov@nexenta.com>
2018-12-05 23:05:42 -08:00
travisn a29b876337 rename ceph v1 crds with the ceph prefix
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-05 16:23:55 -07:00
travisn 62f7e5d6ea ceph: update docs, code, and tests to use v1 crd types
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-05 14:33:11 -07:00
travisn df78f6ad6e remove obsolete rook v1alpha1 types
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-05 14:33:10 -07:00
travisn 501c3a6681 codegen for the ceph v1 types
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-05 14:33:00 -07:00
travisn 364e891ee3 create the ceph v1 types and remove v1alpha1 types
Signed-off-by: travisn <tnielsen@redhat.com>
2018-12-05 14:25:36 -07:00
Yannis Zarkadas 8440d52e7d cassandra: unit tests
Added unit tests for the controller and sidecar packages.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-05 12:48:25 +02:00
Yannis Zarkadas 367d485407 cassandra: sidecar implementation
Implementation of the cassandra sidecar logic.
This sidecar helps us inject our custom values during startup
and provides HTTP liveness and readiness checks.
In the future it will also be used for:
- Shutting Down / Decommissioning a cassandra instance
- Backups
- Restores

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-05 12:47:26 +02:00
Yannis Zarkadas c8c38d8584 cassandra: controller implementation
Control logic for the cassandra controller.
The controller exists in a separate StatefulSet, watches
Kubernetes Objects for changes and acts on those changes to
achieve the desired state.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-05 12:46:33 +02:00
Yannis Zarkadas 0ccc04cccd cassandra: controller scaffolding, image build and manifests
Scaffolding of a Kubernetes controller using the informer-workqueue
pattern.
Dockerfiles and Makefiles to build the cassandra operator image.
Manifests to deploy the Cassandra operator and create a Cluster CRD.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-05 12:44:29 +02:00
Yannis Zarkadas 2b0614a77c cassandra: code generation
Edited codegen.sh to include cassandra.
Then, run make codegen to generate the rest of the files.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-05 12:41:27 +02:00
Yannis Zarkadas bb64ffafe2 cassandra: crd types and boilerplate
API types for Cassandra have been added under /pkg/apis/cassandra.rook.io/types.go
Other files are some necessary boilerplate.

Signed-off-by: Yannis Zarkadas <yanniszarkadas@gmail.com>
2018-12-05 12:41:04 +02:00
Alexander Trost 18b2da5fc5 Merge pull request #2216 from galexrt/fix_2164
Allow specifying custom Ceph user and secret name for mounting
2018-12-04 23:54:27 +01:00
Travis Nielsen 2af0be2cec Merge pull request #2317 from jtlayton/filesystem
Allow filesystem specifications without pools
2018-12-04 15:14:59 -07:00
Alexander Trost ac19f85fbb Allow specifying custom Ceph user and secret name for mounting
Introduce mount security mode for basic multi tenancy

Fixes #2164.

This adds three new parameters/options to StorageClass/flexvolume entry:
* `mountUser`
* `mountSecret`
* `mountSecretNamespace`

Signed-off-by: Alexander Trost <galexrt@googlemail.com>
2018-12-04 20:17:15 +01:00
Blaine Gardner 290b496069 ceph mon/mgr/mds units: verify updated deployments
Verify that the expected deployments are updated in the Ceph mgr,
and mds unit tests.

This also allows those unit tests to pass at all since
UpdateDeploymentAndWait was blocking the unit tests from finishing due
to unexpected behavior of generated `Deployment.Update` unit test mock.

Signed-off-by: Blaine Gardner <blaine.gardner@suse.com>
2018-12-04 11:35:06 -07:00
Blaine Gardner 3d89c78508 ceph: Update deployments on rook img update
Use the `UpdateDeploymentAndWait` method used by the osd operator to
update the deployments for mgrs, mdses, and rgws when the Rook
orchestrator image is updated.

Signed-off-by: Blaine Gardner <blaine.gardner@suse.com>
2018-12-04 11:33:40 -07:00
Jeff Layton 9945e6c91b ceph mds: only permanently remove the fs if rook created it
i.e.: only if there was a data pool specified.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
2018-12-01 20:40:12 -05:00
Jeff Layton dcd9e1e0ff ceph mds: allow filesystem specifications that have no pools
design/external-management.md says that we should allow specifying
Filesystems that have no pools. When this is done, the CRD should
interepret that as a request to just manage the MDS daemons without
creating a new filesystem.

Don't fail the verification if there are no data pools in the spec,
and just skip creating a new filesystem if that's the case.

Also, add a testcase for this usage.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
2018-12-01 20:40:06 -05:00
Jeff Layton 677efb1a37 ceph mds: lift setting of max active MDS out of CreateFilesystem
This seems a bit misplaced, since CreateFilesystem is all about
doing the initial setup of the filesystem, and the daemons are
managed in a different block of code.

In a later patch, we're going to want to allow the CRD to spin
up MDS' for filesystems that already exist. Lift the code that
sets the MDS count into the caller.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
2018-12-01 07:32:46 -05:00
Travis Nielsen 10fa2d9325 Merge pull request #2001 from jcsp/wip-mgr-rbac
Set service account on ceph-mgr pod
2018-12-01 00:58:20 -07:00
travisn ea03f58048 rbac: rename service account rook-ceph-cluster to rook-ceph-osd
Signed-off-by: travisn <tnielsen@redhat.com>
2018-11-30 13:39:58 -07:00
travisn 16c2a56591 add ceph mgr roles to the tests and helm chart
Signed-off-by: travisn <tnielsen@redhat.com>
2018-11-30 13:39:58 -07:00
travisn 69fba00f6b remove --set-uid option from ceph-authtool calls
Newer versions of ceph have deprecated this option, and the commands
will fail if you try to pass it in.

Signed-off-by: travisn <tnielsen@redhat.com>
2018-11-30 13:24:46 -05:00
John Spray cff953696c Set service account on ceph-mgr pod
This is to enable the `rook` ceph-mgr module to drive
the Rook and k8s APIs, exposing functionality via
its own CLI and GUI.

The service account gets access to two different roles:
one for the items that are within the cluster's namespace,
and one for reading configmaps in the rook-ceph-system namespace,
which is needed for reporting available devices.

Signed-off-by: John Spray <john.spray@redhat.com>
2018-11-30 10:50:20 -07:00
Travis Nielsen 9654e51db1 Merge pull request #2290 from travisn/rbd-mirror
Enable rbd mirror daemon in the cluster crd
2018-11-30 08:54:40 -07:00
Alexander Trost 5bc0517bef Merge pull request #2309 from galexrt/pvc_templates_storage_selection
Added PersistentVolumeClaimTemplates to storage selection
2018-11-28 21:46:21 +01:00
Alexander Trost 2aebf5eb26 Added PersistentVolumeClaimTemplates to storage selection
Signed-off-by: Alexander Trost <galexrt@googlemail.com>
2018-11-28 19:12:32 +01:00
travisn 2d1eab4eb2 updated codegen for the object store user
Signed-off-by: travisn <tnielsen@redhat.com>
2018-11-28 09:33:36 -07:00
travisn fdd64bb95a enable rbd mirror daemon in the cluster crd
Signed-off-by: travisn <tnielsen@redhat.com>
2018-11-28 08:29:57 -07:00
Travis Nielsen b239397033 Merge pull request #2293 from travisn/clean-pg-status
Cluster is clean if PGs are scrubbing
2018-11-27 17:31:14 -07:00
Alexander Trost f46efa452c Merge pull request #2282 from galexrt/pera_fix
Use truncated node name for missed sections
2018-11-27 16:30:21 +01:00