When a mon is failed over, the old mon is left in mon quorum
until the new quorum is confirmed in quorum. When host networking
is enabled, this may result in two mons being in quorum with the
same ip, if the new mon is scheduled on the same node as the
down mon. This results in daemon crashes in the cluster since
they are confused by multiple of the same ip in the mon map.
The solution is to avoid scaling down the bad mon when host networking
is enabled. This will ensure the new mon is not scheduled on the
same node as the previous mon. With host networking, either another
node is necessary or the original mon must come back online to
restore quorum.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
currently, we're setting the env `DISCOVER_DAEMON_UDEV_BLACKLIST`
in operator.yaml but we're not reading from the operator env and
setting the value to discover pod env.
Signed-off-by: subhamkrai <srai@redhat.com>
fixes a typo in the 'required' field of the debug-ci input in the
tmate_debug action. this prevented the input from being recognized
correctly by github actions.
Signed-off-by: Oded Viner <oviner@redhat.com>
If the node for an OSD does not exist within kubernetes, the PDB
reconciliation logic will fail to make progress. This can be problematic
as new OSDs added to the cluster may not be protected by OSD PDBs after
they are added since the PDB reconcilation logic is failing.
The solution is to treat non existing OSD nodes as drained which is
logically consistent with treating unschedulable OSD nodes as drained as
you cannot schedule a pod onto a node which does not exist.
Fixes: #16086
Signed-off-by: Elias Carter <elias@dropbox.com>
The create-dev-cluster script used the kvm2 minikube driver on Linux.
On newer Linux flavors like Fedora 41+, the kvm2 driver can have problems
and the qemu2 driver is recommended.
This changes the script to use the qemu2 driver for Linux
Signed-off-by: Michael Adam <obnox@samba.org>
Added a flag in cephcluster CR CSI section
whether the user wants the Rook to manage
the csi users and secrets or create them on own.
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
If RemoveSecret is set remove the secret.
if not set, do below
check if the current cephClient is owner if
secret is present, if not return as its owned
by other controller, if its not present create
or update the secret.
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
When cephClient CR is created by User, the Rook
is creating an User and creating an secret with
its own predefined name + CR name, With this we
are allowing user to specify the secretName
Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
Now rook supports the six most recent minor versions of helm along with
their patch updates. I. e. helm 3.13 and newer is supported.
Fixes: #15980
Co-authored-by: Travis Nielsen <tnielsen@redhat.com>
Signed-off-by: Michael Adam <obnox@samba.org>
When the csi-operator is enabled no need to update
the cm rook-ceph-csi-config configmap.
Otherwise operator will complain about the error
```yaml
ceph-cluster-controller: failed to reconcile CephCluster "rook-ceph/rook-ceph". failed to reconcile cluster "rook-ceph": failed to configure local ceph cluster: failed to create cluster: failed to start ceph monitors: failed to initialize ceph cluster info: failed to save mons: failed to update csi cluster config: waiting for CSI config map to be created: configmaps "rook-ceph-csi-config" not found
```
Signed-off-by: subhamkrai <srai@redhat.com>
Fixes quick disk shredding by using dd to shred data at additional offsets where ceph metadata is duplicated. For full shred, the shred utility remains in use.
Signed-off-by: Vilius Puškunalis <47086537+puskunalis@users.noreply.github.com>
A new helm repo index was created from scratch for every new release
which meant that the release timestamps in the index were basically all
identical.
This change intends to fix that by merging the new index with the
existing one
The first attempt to fix this (PR #16033) did not work properly.
This change is intended to fix the issue for good by closing the gap the
previuous PR has left.
Fixes: #15604
Co-Authored-By: Travis Nielsen <tnielsen@redhat.com>
Signed-off-by: Michael Adam <obnox@samba.org>
skip key generation during reconcile if CephObjectStoreUser is
being deleted to allow cleanup when referenced secret is missing
Signed-off-by: Oded Viner <oviner@redhat.com>
A new helm repo index was created from scratch for every new release
which meant that the release timestamps in the index were basically all
identical.
This change intends to fix that by merging the new index with the
existing one
Fixes: #15604
Signed-off-by: Michael Adam <obnox@samba.org>
Travis Nielsen
Santosh Pillai
subham rai
subhamkrai
Blaine Gardner
dependabot[bot]
Oded Viner
Elias Carter
Deepika Upadhyay
Michael Adam
raaizik
Madhu Rajanna
Muhammad Adil Ghaffar
Vilius Puškunalis
Kyori Sakao