When a PodAntiAffinity is set to mon, detect-version may not be scheduled because detect-version has the same PodAntiAffinity with mon.
This commit fixes this issue.
Closes#5810
Signed-off-by: Hiroshi Muraoka <h.muraoka714@gmail.com>
If there is a typo in the rgw endpoint, then the user will be notified
that the gateway is not accessible or does not return the 200 HTTP code.
Signed-off-by: Sébastien Han <seb@redhat.com>
Most of the CRs expect CephCluster and CephBlockPool were not
reconciling on updates. The predicate must return true on CR diff.
Signed-off-by: Sébastien Han <seb@redhat.com>
adding method to generate secret name in
cephobjectstoreuser status field in order
to programmatically retrieve the secret name
generated by the cephobjectstoreuser crd.
Also added unit test for the method.
Signed-off-by: subhamkrai <subhamkumarrai03@gmail.com>
When hostname has hypen and/or numbers, the mon_endpoints is incorrectly
split and can point to invalid monitor names inside toolbox, Add fix to
split the hostname/ip that can include hypen, underscore or numbers.
Signed-off-by: Vasu Kulkarni <vasu@redhat.com>
Fixes
Signed-off-by: Vasu Kulkarni <vasu@redhat.com>
The output of the bucket policy change should not be logged in case
there is sensitive information in the output.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
Add the ability to provision Ceph OSDs with Drive Groups.
This adds Drive Groups to the CephCluster CRD, and it sets code
in place for propagating this config to the OSD provisioning pod.
Signed-off-by: Blaine Gardner <blaine.gardner@suse.com>
Although the document describes we can't set Pod{,Anti}Affinity
to mon's placement, the current implementation permits to do so.
Signed-off-by: Satoru Takeuchi <satoru.takeuchi@gmail.com>
The version of Jenkins has changed such that setting the official build env
var is no longer comparing correctly. Now we will only assume the env
var is set by the PR to false, therefore the tests need to compare
against false instead of true.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The OSD creation has been fixed in Ceph master, therefore we can
reenable the Ceph mgr test that runs against master.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
moves the admission-controller server to use controller-runtime library for better support, maintainibility of code.
Signed-off-by: Vineet Badrinath <vbadrina@redhat.com>
This commit allows us to configure status check for each daemon:
* "mon": health check on the ceph monitors (quorum)
* "osd": health check on the ceph osds
* "status": ceph health status check
Each check is controlled by the following settings:
* disabled: whether to disable the check (default: false)
* internal: interval to run the check
* timeout: only valid for mons, is the timeout for unresponsive mon
before failling over.
Example to disable the status health check:
```yaml
healthCheck:
daemonHealth:
status:
disabled: true
```
As part of that, pod's livenessprobe can now be configured via the
following settings:
* disabled: whether to enable or not
* probe: override the current probe in place by a new one
```yaml
healthCheck:
livenessProbe:
mon:
disabled: true
```
Closes: https://github.com/rook/rook/issues/5772
Signed-off-by: Sébastien Han <seb@redhat.com>
Adjust the comment for `cephVersion:allowUnsupported` in the
`cluster.yaml` manifest to correctly document that Nautilus and
Octopus are supported (not Mimic and Nautilus).
Signed-off-by: Blaine Gardner <blaine.gardner@suse.com>
Now, when an S3 user gets created, Rook will add the S3 endpoint to the
Secret along with the credentials.
Signed-off-by: Sébastien Han <seb@redhat.com>
For the CephObjectStore, the Status field of the CR has a new property
called "info", which contains a couple of information about the CR.
The first new addition is the endpoint of the object store.
Signed-off-by: Sébastien Han <seb@redhat.com>
Clarify in the docs that changing networking modes after a Cpeh cluster
has been deployed is not supported and can result in a nonfunctional
cluster.
Signed-off-by: Blaine Gardner <blaine.gardner@suse.com>
When mounting a ceph volume, xfs may cause issues with potential
deadlocks if the volume is on the same node as the OSDs. To avoid
this issue the default recommendation is ext4.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The target_size_ratio moved under the parameters element of
the pool settings. While the target size ratio is still supported
under the previous location for backward compatibility, the
example is preferred under parameters.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The OSD daemon has been missing critical flags for logging to stderr
where k8s can capture the logs. Without the --log-to-stderr=true,
all the OSD logging was essentially lost until now.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
Sébastien Han
Jiffin Tony Thottan
Hiroshi Muraoka
Travis Nielsen
subhamkrai
Blaine Gardner
Vasu Kulkarni
Ahmad Nurus S
Satoru Takeuchi
Vineet Badrinath
taeuk_kim