The ceph mons should never be started with an even number in quorum.
The desired state should always be an odd number of mons to ensure
a healthy majority quorum. The operator now rejects a request for
an even number of mons.
If there are not a sufficient number of nodes for mons to be on a
unique host, the operator also returns an early error instead of
getting stuck waiting for a pending mon. This is not foolproof since
there might be fewer available nodes for the mons, but at least it
is a quick check for the common case.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
When a node is not tolerable, also output the node name so it's clear
which not it is which is not tolerable.
Signed-off-by: Thomas Bechtold <tbechtold@suse.com>
When rgw is not responding the aws-sdk retries using exponential
backoff. Having 20 retries means we would wait for a while (hours) to
eventually fail. This is a problem as we cannot reflect the status of
the bucket properly.
Now we use 5 max retries which means around 8 seconds.
Signed-off-by: Sébastien Han <seb@redhat.com>
In cephadm and volume/nfs module, all ganesha daemons within a cluster watch
single config object. Each ganesha daemon in rook watches it's own config
object. This patch updates the ganesha config object name to maintain
uniformity.
Signed-off-by: Varsha Rao <varao@redhat.com>
We can now tune the value of TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES by
setting an annotation on the CephCluster CRD like so:
```
annotation:
osd:
TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES: "134217728"
```
Alternatively, if nothing is specified Rook will read the value of the
file `/etc/sysconfig/ceph`.
Closes: https://github.com/rook/rook/issues/6134
Signed-off-by: Sébastien Han <seb@redhat.com>
Once we close the channel for the monitoring daemons, we update the
monitoring status so as to not call them again.
This is to prevent "close of closed channel" panic.
Co-authored-by: Travis Nielsen <tnielsen@redhat.com>
Signed-off-by: Nitin Goyal <nigoyal@redhat.com>
Signed-off-by: Raghavendra Talur <raghavendra.talur@gmail.com>
on Azure, all SSDs are reporting as a rotational disk.
This commit will add `tuneFastDeviceClass` flag which will
tune for an SSD even if the cloud provider is reporting
rotational. We have a flag `tuneDeviceClass`
which works vice-versa.
Signed-off-by: subhamkrai <srai@redhat.com>
During the initialization encrypted OSD initialization sequence, we check
for the presence of the encrypted container. If it exists we don't try
to open it again since this will result in an error.
However, there is another case we need to handle, when the underlying
device is gone. For instance, if the pod/PV couple was drained and move
back, living the orphan dm. Once the pod comes back, the dm is still
present and perfectly matches. Unfortunately, the underlying disk is
different and thus the dm must be removed and the disk re-opened.
Signed-off-by: Sébastien Han <seb@redhat.com>
During the 1.3 cycle, we had implemented support for OBC external mode
through an Endpoint in the StorageClass parameter. In 1.4, this is not
the case anymore as we directly use the CephObjectStore external mode.
However, we must maintain backward compatibility with the cluster using
the Endpoint and not fail the upgrade.
Thus checking for a CephObjectStore is invalid and should not always be
assumed, now if we detect an Endpoint we return a simple context which
fixes the upgrade.
Signed-off-by: Sébastien Han <seb@redhat.com>
this commit handle golangci-lint linter errcheck.
`errcheck` - Errcheck is a program for checking for
unchecked errors in go programs. These unchecked errors
can be critical bugs in some cases
To see only staticcheck linter output
`golangci-lint run --disable-all -E errcheck`
Signed-off-by: subhamkrai <srai@redhat.com>
On LUKS2 format, the resize operation requires the KEK to be passed. To
maintain a single simple init container for resize, we now open the
encrypted block with "--disable-keyring" so that further operations
won't require the keyring.
Signed-off-by: Sébastien Han <seb@redhat.com>
Ganesha daemons don't need to acess ceph cluster with admin capabilities.
Instead new user with restricted caps can be used.
Signed-off-by: Varsha Rao <varao@redhat.com>
User unlink is performed just before deleting OBC resources.
Since user is internal to OBC and next to remove to delete user,
unlinking user is not necessary.
Fixes: 6300
Signed-off-by: Jiffin Tony Thottan <thottanjiffin@gmail.com>
OSD prepare jobs only need to be run if the OSD deployments on PVCs were
not already started. For efficiency, we now query them in a batch instead
of querying them one-by-one with a label selector, which proves to be very
slow in large clusters.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
The OSD PVCs were being detected one by one with a label selector.
This proves to be slow in large clusters. Instead, we now make this
more efficient by querying for all existing PVCs and looking up their
existing in memory with a map.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
this commit reverts the changes done in
controller file as previous changes failing
in ci and those changes are related to linter
so we can supress them.
Signed-off-by: subhamkrai <srai@redhat.com>
This PR makes the changes in csi templates and
upgrade documentation required for updating
csi sidecar images.
Signed-off-by: Mudit Agarwal <muagarwa@redhat.com>
this commit handle golangci-lint linter staticcheck error.
`staticcheck` - Staticcheck is a go vet on steroids,
applying a ton of static analysis checks.
To see only `staticcheck` linter output
`golangci-lint run --disable-all -E staticcheck`
Signed-off-by: subhamkrai <srai@redhat.com>
The pvc template name has either been require to be empty, or one of the
expected values such as data and metadata. Those who upgrade from 1.1
may find that if they changed it to something else, their upgrade will
fail when not one of these values. If there is only a single template,
always treat it as the data template.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
this commit will enable one more linter ineffassign
in golangci-lint.
This linter throws an error when variable is assigned and never used.
`golangci-lint run --disable-all -E ineffassign` is used detects ineffassign
errors only.
Signed-off-by: subhamkrai <subhamkumarrai03@gmail.com>
The OSD deployments have some env vars duplicated in the pod spec for
the daemon container, which may cause the detection for patch changes
to fail. Now we ensure the env vars are not duplicated in the OSD pod
spec.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
If the pod spec changed, we expect an upgrade to proceed for that daemon.
If the check for a changed pod spec fails, we were skipping the update
of that daemon. Instead of skipping the update, we now assume the pod
spec changed if we fail to detect the change so that we can ensure the upgrade
even if we check for upgrades too often.
Signed-off-by: Travis Nielsen <tnielsen@redhat.com>
Rook is now capable of configuring mirroring between sites. The
implementation works at different levels:
* CephBlockPool: which introduces a new `mirroring` configuration as well
as `statusCheck`. When turned on, Rook will enable mirroring on the
pool. It will also create a bootstrap peer token and store it in a
Kubernetes Secret. The name of that Secret can be found in the Status
field of the CephBlockPool CRD. This token can be fetched and used by
other clusters to configure the site as a peer. Mirroring can be
configured either at the pool or the image level.
* CephRBDMirror: which introduces a new `peers` configuration allowing
Rook to connect to peers by passing a Secret name. The administrator will
create a Kubernetes Secret with 2 keys: 'token' for the bootstrap peer
token and 'pool' for the name of pool. Once detected the rbd-mirror
controller will go ahead and import the peer configuration.
Pool mirroring status example:
```
status:
info:
rbdMirrorBootstrapPeerSecretName: pool-peer-token-test
mirroringInfo:
lastChanged: "2020-09-17T14:47:27Z"
lastChecked: "2020-09-17T14:48:27Z"
summary:
summary:
mode: image
peers:
- client_name: client.rbd-mirror-peer
direction: rx-tx
mirror_uuid: ""
site_name: rhcs
uuid: c50522a4-28a4-4bd3-ba68-e11780308882
site_name: 91eae0dd-06b1-4d2c-91f3-1311c9df382b-rook-ceph
mirroringStatus:
lastChecked: "2020-09-17T14:48:27Z"
summary:
summary:
daemon_health: OK
health: OK
image_health: OK
states:
replaying: 1
```
Signed-off-by: Sébastien Han <seb@redhat.com>
We only need to read the ceph cluster version if the cluster external
has external monitoring enabled.
This avoids any issue during upgrades from 1.3 to 1.4 when monitoring is
enabled and ceph caps have not been updated.
Signed-off-by: Sébastien Han <seb@redhat.com>
Travis Nielsen
Thomas Bechtold
Sébastien Han
Varsha Rao
Ali Maredia
subhamkrai
Raghavendra TalurandTravis Nielsen
subhamkrai
Santosh Pillai
Jiffin Tony Thottan
Madhu Rajanna
Mudit Agarwal
Bhavin Gandhi