the status was still reading from the nil state,
updated the code to dont read from nil object
Signed-off-by: parth-gr <partharora1010@gmail.com>
(cherry picked from commit 0b4f89873c)
LeastUptodateDaemonVersion() unmarshals Ceph's daemon versions JSON
output into `map[string]<type>` maps which are not guaranteed to iterate
in the same order as output by Ceph. Fix this method to no longer assume
iteration order is the same order as Ceph's output.
Signed-off-by: Blaine Gardner <blaine.gardner@ibm.com>
(cherry picked from commit d5c4eca3fb)
The script sel-release-ver.sh was added in a somewhat unnatural place
tests/scripts .
This moves it to a more natural location build/release .
Signed-off-by: Michael Adam <obnox@samba.org>
(cherry picked from commit d7f2a4b4d8)
This script updates examples and docs for a new release.
Invocation in principle: set-release-ver.sh NEW_VER
For example: set-release-ver.sh v1.16.8
Fixes: #15749
Signed-off-by: Michael Adam <obnox@samba.org>
Co-authored-by: Travis Nielsen <tnielsen@redhat.com>
Co-authored-by: Blaine Gardner <b.blaine.gardner@gmail.com>
(cherry picked from commit a1b941f969)
Named containerPort is a useful feature when generating dynamic scrape
targets with prometheus.
This makes it clear that the containerPort is intended for metrics
scraping.
The name matches that which is on the mgr deployment.
Signed-off-by: Justin Cichra <jcichra@cloudflare.com>
(cherry picked from commit 12c36e9472)
ceph-volume raw list will return the main disk instead of the mpath
setup. `ceph-volume lvm zap` on the main disks fails. This PR uses
`lsblk /mnt/<diskreference> ` to fetch the actual mapath disk and use it
for cleaning
Signed-off-by: Santosh Pillai <sapillai@redhat.com>
(cherry picked from commit 1e9da0766e)
The arch specific images were not be pushed since the make
target was missing.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit 63e13868ea)
(cherry picked from commit 9249fb7f91)
The helm charts will now only be published when it is
an officially tagged release build.
The images will only be published to all repos for
dockerhub, quay, and ghcr when it is a tagged release.
The images will be published only to dockerhub for all
master and interim release branch builds.
Remove obsolete makefile option for images.
Ceph is the only image Rook ever expects to build.
Simplify the makefile by removing the legacy option
to select which image to build.
Also included are other small improvements to clean up
the release scripts.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit 7d908d0a5e)
The object store controller was swallowing errors of the
type not found, assuming a certain category of errors
when ceph may still be initializing. But some of these
errors could be user error and they need to troubleshoot
the details of the error, so we need to log the specific
error instead of swallowing it.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit 95a5920bef)
The Ceph upgrade tests have been flaky or reliably failing for various
Ceph/Rook versions due to mon out of space issues. Reinstate the disk
cleanup action, but do not clean up items that rely on the package
manager for removal. Package manager takes 2-4 minutes to operate and is
likely not worth it, even for saving 15+ GB of space.
From testing with this configuration, we can save 11GB of space in about
8 seconds of runtime. This should (hopefully) be plenty of space to keep
total disk usage below 95% and keep mons healthy, and the time needed is
negligible.
Signed-off-by: Blaine Gardner <blaine.gardner@ibm.com>
(cherry picked from commit e97103d859)
In case of multicluster service, we need to always delete the mon-canary
pods even if there is any failure while exporting the service, else
multiple mon canary pods will get created.
Signed-off-by: Santosh Pillai <sapillai@redhat.com>
(cherry picked from commit 75c6bb8cd2)
The upgrade tests in master have been upgrading from 1.14 to
master. In anticipation of the v1.16 release, we change
the upgrade tests to start from v1.15 to test if there
are any regressions in the supported upgrades.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit 506407b5ea)
The tests sometimes see that the mons run out of space
in the CI, therefore, we reduce the space required
from 30% down to 10%.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit ca1d2ed3e3)
Update the examples and docs and operator base image to
the latest ceph patch release v19.2.2.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit f62e6d9cc2)
Ceph uses uint64 (unsigned long) type for CephFS GID. Rook was using
int, which was causing failures when the GID overflowed Rook's type.
Update Rook's type to match Ceph's.
Signed-off-by: Blaine Gardner <blaine.gardner@ibm.com>
(cherry picked from commit 5b27189ac4)
In some cases, for example, kube-api nods, CephBlockPool's status
could be failed to update. In that case, rook-operator does not
tries to update status again. This fix requests for requeue if pool's
status update is failed.
Signed-off-by: Peter Razumovsky <prazumovsky@mirantis.com>
(cherry picked from commit d7c67fecee)
there was very edge case,a nd that result in nil
point exception
During osd upgrades if mirror health check is called
the mirror status cmd doesnt respond
So its important to check if the value exist
before getting its summary
Signed-off-by: parth-gr <partharora1010@gmail.com>
(cherry picked from commit 0c5d3ee3a5)
The default object store ingress port will be set to the
securePort (if set), or the gateway port field. To override
that default, a new port field can be set on the ingress.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit 5ad38cb579)
The networking.k8s.io/v1beta1 has been obsolete since
K8s 1.22, long out of the supported versions, so we
can remove those legacy checks and require the v1
networking api.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
(cherry picked from commit c40477ccc2)
the maxunavailable count was getting reset back to 1 after 60 seconds.
This PR prevents resetting of the maxUnavailable OSD count.
Signed-off-by: Santosh Pillai <sapillai@redhat.com>
(cherry picked from commit e83ca35531)
Travis Nielsen
parth-gr
Blaine Gardner
Artem Torubarov
Michael Adam
mergify[bot]
Justin Cichra
Santosh Pillai
subham rai
Bryan Jones
Peter Razumovsky