Commit Graph
7932 Commits
Author SHA1 Message Date
Sébastien Han decc481d40 Merge pull request #8740 from leseb/examples-layout
core: change directory layout
2021-11-30 16:02:37 +01:00
Sébastien Han c890710b63 core: change directory layout
As per discussion, proposing a new layout for the charts/yaml/olm files.

./deploy
├── charts
│   ├── rook-ceph
│   │   └── templates
│   └── rook-ceph-cluster
│       └── templates
├── examples
│   ├── csi
│   │   ├── cephfs
│   │   └── rbd
│   ├── flex
│   ├── monitoring
│   ├── pre-k8s-1.16
└── olm
    └── assemble

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-30 09:12:53 +01:00
Sébastien Han 83f7c2be87 Merge pull request #9259 from parth-gr/deviceClass
osd: update existing OSDs with deviceClass
2021-11-29 18:06:08 +01:00
parth-gr cbe505d122 osd: update existing OSDs with deviceClass
If we apply useAllNodes to false for the current deployment,
the OSDs should get updated with the individual nodes values and config,
The deviceClass was not updating to the existing OSDs because there was
bug in the check.
The check osdInfo.DeviceClass == "" which should be
checked like this osdInfo.DeviceClass == "None"

Updated the code so OSDs can make use of the devices present

Signed-off-by: parth-gr <paarora@redhat.com>
2021-11-29 21:40:02 +05:30
Sébastien Han bf662ec7b9 Merge pull request #9264 from leseb/fix-9151
cephfs-mirror: various fixes for random bootstrap peer import errors
2021-11-29 15:45:14 +01:00
Sébastien Han bd962e3e85 Merge pull request #9104 from BlaineEXE/nfs-restart-with-configmap
nfs: restart nfs servers when configmap is updated
2021-11-26 17:06:16 +01:00
Sébastien Han d8a8b05c1f cephfs-mirror: use combined output to possibly catch peer import error
By adding a combined output to the executor we might be able to fetch more
error messages.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-26 15:53:30 +01:00
Sébastien Han 7f7a72d941 core: add the ability to execute ceph commands with a combined output
Sometimes Ceph uses a different standard output to return errors or
merges standard error to standard out. So let's allow some commands to
return both in the output.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-26 15:52:49 +01:00
Sébastien Han d1cdba420a cephfs-mirror: try to mitigate peer import error
Some users have reported issues while adding the token, this is not
always reproducable so perhaps it's a typo when importing the token and
adding trailing spaces.

Closes: https://github.com/rook/rook/issues/9151
Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-26 15:45:01 +01:00
Sébastien Han 2519018b26 Merge pull request #9262 from leseb/bot-warn
bot: add bot message when pushing on a release branch
2021-11-26 10:06:33 +01:00
Sébastien Han ae6521f277 Merge pull request #9261 from leseb/master-for-9249
object: fix rgw ceph config
2021-11-26 10:02:59 +01:00
Sébastien Han c2d36c3343 bot: add bot message when pushing on a release branch
Normally, only the mergify bot should send out PR to the release
branches. All contributions should generally go in master first then be
backported. Let's warn to avoid merging PR send against a release
branch.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-25 18:51:38 +01:00
Olivier 561cede1ef object: fix rgw ceph config
use Zone and ZoneGroup instead of storename for rgw_zone and rgw_zonegroup

Signed-off-by: Olivier Bouffet <olivier.bouffet@infomaniak.com>
(cherry picked from commit c92270cd66)
2021-11-25 18:38:42 +01:00
Sébastien Han 0dcfefa3a1 Merge pull request #9258 from leseb/revert-snyk
build: revert "build: images/cross/Dockerfile to reduce vulnerabilities"
2021-11-25 15:59:42 +01:00
Sébastien Han 619f5eee6e build: revert "build: images/cross/Dockerfile to reduce vulnerabilities"
This reverts commit ceec0bc48c. The build
is failing to push images to master. This is not a critical update so
let's remove.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-25 14:53:21 +01:00
Sébastien Han 35b28f760f Merge pull request #9250 from rook/snyk-fix-8a1398b3ebbfb4eeed7c40de45c7b73b
[Snyk] Security upgrade ubuntu from xenial-20210114 to impish-20211015
2021-11-25 13:49:46 +01:00
snyk-bot ceec0bc48c build: images/cross/Dockerfile to reduce vulnerabilities
The following vulnerabilities are fixed with an upgrade:
- https://snyk.io/vuln/SNYK-UBUNTU1604-LIBGCRYPT20-1585790
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-25 11:59:22 +01:00
Sébastien Han 7a223adde4 Merge pull request #9230 from leseb/osd-rm-check
osd: check if osd is safe-to-destroy before removal
2021-11-25 10:55:34 +01:00
Sébastien Han ad2c3c2ae6 Merge pull request #8931 from BlaineEXE/test-rgw-multisite-in-nightly-tests
test: test rgw multisite nightly
2021-11-25 10:25:57 +01:00
Sébastien Han 7402c2cce6 osd: check if osd is ok-to-stop before removal
If multiple removal jobs are fired in parallel, there is a risk of
losing data since we will forcefully remove the OSD. It's also simply
true if a single OSD is not safe to destroy, there is also a risk of
data loss.

So now, we check if the OSD is safe-to-destroy first and then proceed.
The code waits forever and retries every minute unless the
--force-osd-removal flag is passed.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-25 10:21:32 +01:00
Sébastien Han 3e299fbe13 Merge pull request #9236 from leseb/fix-9234
core: fix openshift security context
2021-11-24 18:27:29 +01:00
Travis Nielsen 68f8d5c217 Merge pull request #9235 from rook/snyk-fix-764e1f7e4c3e1bb5ec7e16c52097dba6
[Snyk] Security upgrade ubuntu from 16.04 to xenial-20210114
2021-11-24 10:19:09 -07:00
Travis Nielsen 9d3bcb4a78 Merge pull request #9243 from LittleFox94/9241-update-prometheus
monitoring: update label on prometheus resources
2021-11-24 10:16:45 -07:00
Mara Sophie Grosch 733878b1eb monitoring: update label on prometheus resources
Updating the promethes reources (PrometheusRule and ServiceMonitor) is
done by fetching the current resource from the server and updating the
spec on it. This commit makes it also apply the labels, so users can
update them via rook CRDs.

Closes: https://github.com/rook/rook/issues/9241
Signed-off-by: Mara Sophie Grosch <littlefox@lf-net.org>
2021-11-24 17:17:02 +01:00
snyk-bot 38b2af711a build: images/cross/Dockerfile to reduce vulnerabilities
The following vulnerabilities are fixed with an upgrade:
- https://snyk.io/vuln/SNYK-UBUNTU1604-LIBGCRYPT20-1585790
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131
- https://snyk.io/vuln/SNYK-UBUNTU1604-SYSTEMD-1320131

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-24 15:38:15 +01:00
Sébastien Han e008464327 Merge pull request #9224 from leseb/fix-9205
nfs: only set the pool size when it exists and always run default pool creation
2021-11-24 14:36:14 +01:00
Sébastien Han f3142847d3 nfs: always run default pool creation
Previously, if the pool was present we would not run the pool creation
again. This is a problem if the pool spec changes, the new settings will
never be applied.

Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-24 14:04:19 +01:00
Sébastien Han b38f430c26 core: fix openshift security context
The MKNOD capability was missing and due to recent addition some pod now
only require this cap as well as privileged.
The cap must be explicitly exposed so it can be requested by a pod.

Closes: https://github.com/rook/rook/issues/9234
Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-24 11:22:47 +01:00
Sébastien Han 12638cb56f Merge pull request #9233 from arugal/typo
helm: update comments to match the default image in the code
2021-11-24 09:28:37 +01:00
zhangwei d63d3cf0c4 helm: update comments to match the default image in the code
update comments to match the default image in the code.

Signed-off-by: zhangwei <zhangwei24@apache.org>
2021-11-24 11:07:43 +08:00
Travis Nielsen ba54567e8f Merge pull request #9176 from TomHellier/9174-ingress-support-more-k8s-versions
helm: Allow further configurability of the ingress version
2021-11-23 13:47:33 -07:00
Travis Nielsen d340bccd44 Merge pull request #9202 from thotz/hpakedadoc
docs: add details about HPA via KEDA
2021-11-23 10:42:03 -07:00
Travis Nielsen 801b5d65a7 Merge pull request #9180 from LittleFox94/8502-monitoring-labels
#8502: Make monitoring object labels overridable
2021-11-23 09:44:06 -07:00
Sébastien Han f1c896af72 Merge pull request #9226 from leseb/fix-9166
docs: move the test readme to the official doc
2021-11-23 16:36:47 +01:00
Sébastien Han f9f08c88ca docs: move the test readme to the official doc
The instructions to deploy a developer environment are useful and should
be part of the official documentation instead of being hidden in the
repo's README.md. Also, remove ancient documentation to deploy with
kubespray a multi-node env.
Also, add a warning to always run tests env in a virtual machine.

Closes: https://github.com/rook/rook/issues/9166
Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-23 15:56:28 +01:00
Jiffin Tony Thottan d32c837e35 docs: add details about HPA via KEDA
By default HPA can use details about memory or CPU consumption for
autoscaling, but also it can use custom metrics as well. There are alot
provides supports HPA via customer and one of them is KEDA project. Here
it is done with help of Prometheus Scaler

Signed-off-by: Jiffin Tony Thottan <thottanjiffin@gmail.com>
2021-11-23 16:50:18 +05:30
Sébastien Han c3accdcc3a nfs: only set the pool size when it exists
For CRD not using the new nfs spec that includes the pool settings,
applying the "size" property won't work since it is set to 0. The pool
still gets created but returns an error. The loop is re-queued but on
the second run the pool is detected so no further configuration is done.

Closes: https://github.com/rook/rook/issues/9205
Signed-off-by: Sébastien Han <seb@redhat.com>
2021-11-23 11:07:43 +01:00
Sébastien Han e9f9a40138 Merge pull request #9212 from travisn/admin-test-cluster
core: Ensure cluster name is available on cluster info
2021-11-22 11:50:55 +01:00
Tom Hellier ba44602477 helm: allow further configurability of ingress version
The ingress api version changed when it went to v1, and this has caused some upheaval
throughout the kubernetes ecosystem. This commit uses a common method of deciding which
ingress api to use, and allows the optional override of the kubernetes version
presented to helm using the helm build-in capabilities.
also add an ingress into the helm integration tests so any regressions to how ingresses
are handled in the future are caught easier.

Closes rook#9174

Signed-off-by: Tom Hellier <me@tomhellier.com>
2021-11-22 10:03:02 +00:00
Blaine Gardner fcd0d90c3a Merge pull request #9094 from BlaineEXE/clean-up-resource-health-checkers-when-removing-finalizers
pool: file: object: clean up health checkers for both types of deletion
2021-11-19 11:57:25 -07:00
Blaine Gardner 03ba7dec64 pool: file: object: clean up stop health checkers
Clean up the code used to stop health checkers for all controllers
(pool, file, object). Health checkers should now be stopped when
removing the finalizer for a forced deletion when the CephCluster does
not exist. This prevents leaking a running health checker for a resource
that is going to be imminently removed.

Also tidy the health checker stopping code so that it is similar for all
3 controllers. Of note, the object controller now uses namespace and
name for the object health checker, which would create a problem for
users who create a CephObjectStore with the same name in different
namespaces.

Signed-off-by: Blaine Gardner <blaine.gardner@redhat.com>
2021-11-19 10:29:12 -07:00
Blaine Gardner e80368674d test: run RGW multisite test in nightly job
In the nightly job, run the test with the latest Ceph version so we can
detect if there are RGW changes in Ceph that might break multisite. Use
a reusable GitHub action workflow to duplicate as little code as
possible.

Signed-off-by: Blaine Gardner <blaine.gardner@redhat.com>
2021-11-19 09:56:41 -07:00
Travis Nielsen 1afd322650 core: ensure cluster name is available on cluster info
The cluster info is important context for the cluster controller to
create the cluster, and all the fields must be properly set.
A test cluster name was being set temporarily, resulting in
mons incorrectly getting the wrong cluster CR name. There is no
known issue from the temporary value, it was just exposed by
https://github.com/rook/rook/pull/8678 setting the value to a label.

Now the functions are more clearly named so only unit and
integration tests should be using the test value for the cluster
name where it is not important.

Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
2021-11-18 15:30:26 -07:00
Blaine Gardner b9faa9d82f Merge pull request #9187 from BlaineEXE/bluestore_fsck_quick_fix_on_mount-upgrade-warning
docs: add OMAP quick fix warning to upgrade guide
2021-11-18 12:47:11 -07:00
Blaine Gardner cd832e5832 Merge pull request #8579 from thotz/vaultsslsupport
ceph: add support in RGW to communicate vault with TLS
2021-11-18 12:32:38 -07:00
Travis Nielsen 3367f7e30b Merge pull request #9203 from parth-gr/mon-clusterInfo-cr
mon: set cluster name to mon cluster
2021-11-18 10:57:08 -07:00
Blaine Gardner 1f8719650a docs: add OMAP quick fix warning to upgrade guide
Ceph has recently reported that it may be unsafe to upgrade clusters
from Nautilus/Octopus to Pacific v16.2.0 through v16.2.6. We are
tracking this in Rook issue https://github.com/rook/rook/issues/9185.
Add a warning to the upgrade doc about this.

Signed-off-by: Blaine Gardner <blaine.gardner@redhat.com>
2021-11-18 10:04:25 -07:00
parth-gr aea2856566 mon: set cluster name to mon cluster
the mon cluster clusterInfo is intiated seprately,
and misses out to set the cluster name and use default name as testing
from AdminClusterInfo.

Part-of: https://github.com/rook/rook/issues/9159
Signed-off-by: parth-gr <paarora@redhat.com>
2021-11-18 20:28:51 +05:30
Blaine Gardner d62abfd98d Merge pull request #9201 from BlaineEXE/move-rbd-csi-nodeplugin-clusterrolebinding-in-charts
helm: move rbd-csi-nodeplugin from PSP resources
2021-11-17 17:43:28 -07:00
Blaine Gardner dd85bd2d33 helm: move rbd-csi-nodeplugin from PSP resources
Move the rbd-csi-nodeplugin ClusterRoleBinding from the chart PSP
resources to its proper place with the rest of the operator-wide
ClusterRoleBindings.

Signed-off-by: Blaine Gardner <blaine.gardner@redhat.com>
2021-11-17 16:34:54 -07:00