RGW can only serve a single certificate. This limitation means that the
prior behavior of using the default service for admin ops when TLS is
enabled may mean it requires additional complex certificate management
to make sure the object store uses a certificate valid for Rook internal
admin ops and user connections.
This is needlessly complex for users. Instead, change Rook's behavior
and documentation to clarify that it will use the same endpoint intended
for S3 client applications. This means that users have a more
straightforward path to enabling both Rook and consuming applications.
More info: https://github.com/rook/rook/issues/14530
Signed-off-by: Blaine Gardner <blaine.gardner@ibm.com>
(cherry picked from commit b4a2285aa6)
issue:
external-provisioner (v5) enabled topology feature-gate by default
and the current implementation in Rook uses a conditional block to
enable the topology feature gate. This approach now does not directly
reflect the state of the `CSI_ENABLE_TOPOLOGY`.
fix:
replacing the conditional block with a direct use of the
`CSI_ENABLE_TOPOLOGY` for flag value.
Signed-off-by: Praveen M <m.praveen@ibm.com>
(cherry picked from commit e30e5e881f)
Add CephObjectStore spec.hosting.advertiseEndpoint configuration. This
provides a clear documented default for which endpoint Rook "advertises"
to dependent resources like CephObjectStores, OBCs, and COSI
Buckets/Accesses and allows users to override the default behavior if
desired.
The current default is to round-robin an endpoint from
spec.hosting.dnsNames, which has proven to be troublesome for some
users' object store configurations. This change provides much-needed
disambiguation for users.
This may be a breaking change for some existing spec.hosting.dnsNames
users. This is unexpected but is documented.
Signed-off-by: Blaine Gardner <blaine.gardner@ibm.com>
(cherry picked from commit a2b0b6449c)
Fixes: #14518
A recent change seems to have broken the permissions of the auto-assign
action.
This tries to fix this by making permissions more specific.
Signed-off-by: Michael Adam <obnox@samba.org>
Create new topic in the troubleshooting guide to explain how to fix a new rook-ceph deployment that could remain in an unhealthy state or fails to config because of the LimitNOFILE value specified in the containerd.service configuration.
Signed-off-by: Nicola Fantacuzzi <nfantacuzzi@gmail.com>
The daily upgrade tests were checking for a specific
version of the rook image instead of the local_build tag
that is set as the rook base image since 14486.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
If the dashboard cert already exists when ssl is enabled,
we need to not restart the dashboard module. The module
should only be restarted when there is a change to the
dashboard configuration. When ssl is enabled, the dashboard
is essentially being restarted every time there is a reconcile
which is causing issues with the mgr getting blocklisted
unnecessarily by ceph during failover.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The helm upgrade tests have been failing frequently, but not
always, on the oldest version of K8s that is tested in the CI
for the past few months. Add a retry to attempt to get
the CI passing more consistently.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The upgrade tests in master have been upgrading from 1.13 to
master. In anticipation of the v1.15 release, we change
the upgrade tests to start from v1.14 to test if there
are any regressions in the supported upgrades.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The daily ceph upgrade tests were running on Rook v1.13.
This was causing the squid upgrade tests to fail since
1.13 does not support Squid. The purpose of the upgrade
tests is to test the ceph upgrades, therefore the daily
tests will just test them based on rook master.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
currently the output in the cm was not human readable
and can not easily used to create a config.ini file
Also for now removed the support of printing arg in upstream
Signed-off-by: parth-gr <partharora1010@gmail.com>
Fixes: #14466
The OpenSSF scorecard report
https://scorecard.dev/viewer/?uri=github.com/rook/rook
warns about excessive token-permissions in the rook project's github CI
workflows.
This bulk change aims at improving this by
1. adding top-level read permissions where missing for integration tests.
2. adding write permissions as needed for the auto-assign workflow.
3. removing the warned-about and unused create-tag workflow
Signed-off-by: Michael Adam <obnox@samba.org>
Co-authored-by: Blaine Gardner <b.blaine.gardner@gmail.com>
Normally the device class of an OSD is determined at provisioning
and is not updated thereafter. In some scenarios the admin may
want to force update the device class to a new value. The device
classes can be updated by first setting allowDeviceClassUpdate
in the storage spec of the cephcluster, then updating the
device class specified on the deviceSets or other OSDs.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
currently, it was blocking both the mds ip
instead its should block ip of of mds which is in
the same node which is down.
Signed-off-by: subhamkrai <srai@redhat.com>
Blaine Gardner
mergify[bot]
Praveen M
Travis Nielsen
Michael Adam
sp98
subham rai
dependabot[bot]
nicofnt
parth-gr
Artem Torubarov
yingshanghuangqiao
Satoru Takeuchi
subhamkrai