Ceph
PLEASE NOTE: This document applies to an unreleased version of Rook. It is strongly recommended that you only use official releases of Rook, as unreleased versions are subject to changes and incompatibilities that will not be supported in the official releases.
If you are using an official release version of Rook, you should refer to the documentation for your specific version.
Documentation for other releases can be found by using the version selector in the bottom left of any doc page.Ceph NFS Server CRD
Overview
Rook allows exporting NFS shares of a CephFilesystem or CephObjectStore through the CephNFS custom resource definition. This will spin up a cluster of NFS Ganesha servers that coordinate with one another via shared RADOS objects. The servers will be configured for NFSv4.1+ access only, as serving earlier protocols can inhibit responsiveness after a server restart.
WARNING: We do not recommend using NFS in Ceph v16.2.0 through v16.2.6. If you are using Ceph v15, we encourage you to upgrade directly to Ceph Pacific v16.2.7. Upgrade steps are outlined below.
Samples
The following sample assumes Ceph v16 and will create a two-node active-active cluster of NFS Ganesha gateways.
apiVersion: ceph.rook.io/v1
kind: CephNFS
metadata:
name: my-nfs
namespace: rook-ceph
spec:
# For Ceph v15, the rados block is required. It is ignored for Ceph v16.
rados:
# RADOS pool where NFS configs are stored.
# In this example the data pool for the "myfs" filesystem is used.
# If using the object store example, the data pool would be "my-store.rgw.buckets.data".
# Note that this has nothing to do with where exported file systems or object stores live.
pool: myfs-data0
# RADOS namespace where NFS client recovery data is stored in the pool.
namespace: nfs-ns
# Settings for the NFS server
server:
# the number of active NFS servers
active: 2
# A key/value list of annotations
annotations:
# key: value
# where to run the NFS server
placement:
# nodeAffinity:
# requiredDuringSchedulingIgnoredDuringExecution:
# nodeSelectorTerms:
# - matchExpressions:
# - key: role
# operator: In
# values:
# - mds-node
# tolerations:
# - key: mds-node
# operator: Exists
# podAffinity:
# podAntiAffinity:
# topologySpreadConstraints:
# The requests and limits set here allow the ganesha pod(s) to use half of one CPU core and 1 gigabyte of memory
resources:
# limits:
# cpu: "500m"
# memory: "1024Mi"
# requests:
# cpu: "500m"
# memory: "1024Mi"
# the priority class to set to influence the scheduler's pod preemption
priorityClassName:
NFS Settings
RADOS Settings
NFS configuration is stored in a Ceph pool so that it is highly available and protected. How that is
configured changes depending on the Ceph version. Configuring the pool is done via the rados
config.
WARNING: Do not use erasure coded (EC) pools for NFS. NFS-Ganesha uses OMAP which is not supported by Ceph’s erasure coding.
For Ceph v16 or newer
poolConfig
: (optional) The pool settings to use for the RADOS pool. It matches the CephBlockPool specification. The settings will be applied to a pool named.nfs
on Ceph v16.2.7 or newer.
For Ceph v15
pool
: (mandatory) The Ceph pool where NFS configuration is stored.namespace
: (mandatory) The namespace in thepool
where configuration objects will be stored.
Rook ignores both pool
and namespace
(see above) settings when running Ceph v16
or newer.
Creating Exports
When a CephNFS is first created, all NFS daemons within the CephNFS cluster will share a configuration with no exports defined.
For Ceph v16 or newer
For Ceph v16.2.0 through v16.2.6, exports cannot be managed through the Ceph dashboard, and newly-created Ceph command line tools are lacking. We highly recommend using Ceph v16.2.7 or higher with NFS, which fixes bugs and streamlines export management, allowing exports to be created via the Ceph Dashboard and the Ceph CLI. With v16.2.7 or higher, the Ceph dashboard and Ceph CLI will be able to manage the same NFS exports interchangeably as desired.
Using the Ceph Dashboard
Exports can be created via the Ceph dashboard for Ceph v16 as well. To enable and use the Ceph dashboard in Rook, see here.
Using the Ceph CLI
The Ceph CLI can be used from the Rook toolbox pod to create and manage NFS exports. To do so, first ensure the necessary Ceph mgr modules are enabled and that the Ceph orchestrator backend is set to Rook.
ceph mgr module enable rook
ceph mgr module enable nfs
ceph orch set backend rook
Ceph’s NFS CLI can create NFS exports
that are backed by CephFS (a CephFilesystem) or
Ceph Object Gateway (a CephObjectStore).
cluster_id
or cluster-name
in the Ceph NFS docs normally refers to the name of the NFS cluster,
which is the CephNFS name in the Rook context.
For creating an NFS export for the CephNFS and CephFilesystem example manifests, the below command
can be used. This creates an export for the /test
pseudo path.
ceph nfs export create cephfs my-nfs /test myfs
The below command will list the current NFS exports for the example CephNFS cluster, which will give the output shown for the current example.
ceph nfs export ls my-nfs
[
"/test"
]
The simple /test
export’s info can be listed as well. Notice from the example that only NFS
protocol v4 via TCP is supported.
ceph nfs export info my-nfs /test
{
"export_id": 1,
"path": "/",
"cluster_id": "my-nfs",
"pseudo": "/test",
"access_type": "RW",
"squash": "none",
"security_label": true,
"protocols": [
4
],
"transports": [
"TCP"
],
"fsal": {
"name": "CEPH",
"user_id": "nfs.my-nfs.1",
"fs_name": "myfs"
},
"clients": []
}
If you are done managing NFS exports and don’t need the Ceph orchestrator module enabled for anything else, it may be preferable to disable the Rook and NFS mgr modules to free up a small amount of RAM in the Ceph mgr Pod.
ceph mgr module disable nfs
ceph mgr module disable rook
Mounting exports
Each CephNFS server has a unique Kubernetes Service. This is because NFS clients can’t readily
handle NFS failover. CephNFS services are named with the pattern
rook-ceph-nfs-<cephnfs-name>-<id>
<id>
is a unique letter ID (e.g., a, b, c, etc.) for a given
NFS server. For example, rook-ceph-nfs-my-nfs-a
.
For each NFS client, choose an NFS service to use for the connection. With NFS v4, you can mount all exports at once to a mount location.
mount -t nfs4 -o proto=tcp <nfs-service-ip>:/ <mount-location>
For Ceph v15
Exports can be created via the Ceph dashboard for Ceph v15. To enable and use the Ceph dashboard in Rook, see here.
Enable the creation of NFS exports in the dashboard for a given cephfs or object gateway pool by running the following command in the toolbox container:
- For a single CephNFS cluster
ceph dashboard set-ganesha-clusters-rados-pool-namespace <pool>[/<namespace>]
- For multiple CephNFS clusters
ceph dashboard set-ganesha-clusters-rados-pool-namespace <cephnfs-name>:<pool>[/<namespace>](,<cephnfs-name>:<pool>[/<namespace>])*
For each of the multiple entries above,
cephnfs-name
is the name given to CephNFS resource by the manifest’smetadata.name
:my-nfs
for the example earlier in this document.pool
andnamespace
are the same configured via the CephNFS spec’srados
block.
You should now be able to create exports from the Ceph dashboard.
You may need to enable exports created by the dashboard before they will work!
Creating exports via the dashboard does not necessarily enable them. Newer versions of Ceph v15 enable the exports automatically, but not all. To ensure the exports are created automatically, use Ceph v15.2.15 or higher. Otherwise, you must take the manual steps below to ensure the exports are enabled.
To enable exports, we are going to modify the Ceph RADOS object (stored in Ceph) that defines the configuration shared by all NFS daemons.
Please note that <pool>
and <namespace>
will continue to refer to the configured rados
spec’s
pool
and namespace
for a particular CephNFS cluster.
List the shared configuration objects in a Ceph pool with this command from the Ceph toolbox.
rados --pool <pool> --namespace <namespace> ls
The output may look something like below after you have created two exports. Here we have used the
my-nfs
example CephNFS.
conf-nfs.my-nfs
export-1
export-2
grace
rec-0000000000000002:my-nfs.a
The configuration of NFS daemons, and enabling exports, is controlled by the conf-nfs.my-nfs
object in this example. The object name follows the conf-nfs.<cephnfs-name>
pattern.
Get the contents of the config file, which may be empty as in this example.
rados --pool <pool> --namespace <namespace> get conf-nfs.my-nfs my-nfs.conf
cat my-nfs.conf
Modify the my-nfs.conf
file above to add URLs for enabling exports.
%url "rados://<pool>/<namespace>/export-1"
%url "rados://<pool>/<namespace>/export-2"
Then write the modified file to the RADOS config object.
rados --pool <pool> --namespace <namespace> put conf-nfs.my-nfs my-nfs.conf
Verify the changes are saved by getting the config again, just as before.
rados --pool <pool> --namespace <namespace> get conf-nfs.my-nfs my-nfs.conf
cat my-nfs.conf
%url "rados://<pool>/<namespace>/export-1"
%url "rados://<pool>/<namespace>/export-2"
Upgrading from Ceph v15 to v16
We do not recommend using NFS in Ceph v16.2.0 through v16.2.6 due to bugs in Ceph’s NFS implementation. If you are using Ceph v15, we encourage you to upgrade directly to Ceph v16.2.7.
Prep
To upgrade, first follow the usual Ceph upgrade steps. When the upgrade completes, this will result in NFS exports that no longer work. The dashboard’s NFS management will also be broken. We must now migrate the NFS exports to Ceph’s new management method.
We will do all work from the toolbox pod. Exec into an interactive session there.
First, unset the previous dashboard configuration with the below command.
ceph dashboard set-ganesha-clusters-rados-pool-namespace ""
Also ensure the necessary Ceph mgr modules are enabled and that the Ceph orchestrator backend is set to Rook.
ceph mgr module enable rook
ceph mgr module enable nfs
ceph orch set backend rook
Step 1
Pick a CephNFS to work with and make a note of the spec.rados.pool
and spec.rados.namespace
. If
the pool
is not set, it is .nfs
. We will refer to these as pool/<pool>
and
namespace/<namespace>
for the remainder of the steps. Also note the name of the CephNFS resource,
which will be referred to as CephNFS name or <cephnfs-name>
.
Step 2
List the exports defined in the pool.
rados --pool <pool> --namespace <namespace> ls
This may look something like below.
grace
rec-0000000000000002:my-nfs.a
export-1
export-2
conf-nfs.my-nfs
Step 3
For each export above, save the export to an <export>.conf
file.
EXPORT="export-1" # "export-2", "export-3", etc.
rados --pool <pool> --namespace <namespace> get "$EXPORT" "/tmp/$EXPORT.conf"
The file should contain content similar to what is shown here.
$ cat /tmp/export-1.conf
EXPORT {
export_id = 1;
path = "/";
pseudo = "/test";
access_type = "RW";
squash = "no_root_squash";
protocols = 4;
transports = "TCP";
FSAL {
name = "CEPH";
user_id = "admin";
filesystem = "myfs";
secret_access_key = "AQAyr69hwddJERAAE9WdFCmY10fqehzK3kabFw==";
}
}
Step 4
We will now import each export into Ceph’s new format. Perform this step for each export you wish to migrate.
First remove the FSAL
configuration block’s user_id
and secret_access_key
configuration items.
It is sufficient to delete the lines in the /tmp/<export>.conf
file using vi
or some other
editor. The file should look similar to below when the edit is finished.
$ cat /tmp/<export>.conf
EXPORT {
export_id = 1;
path = "/";
pseudo = "/test";
access_type = "RW";
squash = "no_root_squash";
protocols = 4;
transports = "TCP";
FSAL {
name = "CEPH";
filesystem = "myfs";
}
}
Now that the old user and access key are removed, import the export. There should be no errors, but if there are, follow the error message instructions to proceed.
ceph nfs export apply <cephnfs-name> -i /tmp/<export>.conf
Step 5
Once all exports have been migrated for the current CephNFS, it is good to verify the exports. Use
ceph nfs export ls <cephnfs-name>
to list all exports (identified by the pseudo path), and use
ceph nfs export info <cephnfs-name> <export-pseudo>
to inspect the configuration. An export
configuration may look something like below. The v16 CLI section above shows
this in more detail.
Step 6
Repeat these steps for each other CephNFS.
Clean up all <export>.conf
files before moving onto subsequent CephNFSes to avoid confusion.
rm -f /tmp/export-*.conf
Wrap-up
Once you are finished migrating all CephNFSes, the migration is complete. If you wish to use the Ceph dashboard to manage exports, you should now be able to find them all listed there.
If you are done managing NFS exports via the CLI and don’t need the Ceph orchestrator module enabled for anything else, it may be preferable to disable the Rook and NFS mgr modules to free up a small amount of RAM in the Ceph mgr Pod.
ceph mgr module disable nfs
ceph mgr module disable rook
Scaling the active server count
It is possible to scale the size of the cluster up or down by modifying the spec.server.active
field. Scaling the cluster size up can be done at will. Once the new server comes up, clients can be
assigned to it immediately.
The CRD always eliminates the highest index servers first, in reverse order from how they were started. Scaling down the cluster requires that clients be migrated from servers that will be eliminated to others. That process is currently a manual one and should be performed before reducing the size of the cluster.
Advanced configuration
All CephNFS daemons are configured using shared configuration objects stored in Ceph. In general, users should only need to modify the configuration object. Exports can be created via the simpler Ceph-provided means documented above.
For configuration and advanced usage, the format for these objects is documented in the NFS Ganesha project.
Use Ceph’s rados
tool from the toolbox to interact with the configuration object. The below
command will get you started by dumping the contents of the config object to stdout. The output may
look something like the example shown.
rados --pool <pool> --namespace <namespace> get conf-nfs.<cephnfs-name> -
%url "rados://<pool>/<namespace>/export-1"
%url "rados://<pool>/<namespace>/export-2"
rados ls
and rados put
are other commands you will want to work with the other shared
configuration objects.
Of note, it is possible to pre-populate the NFS configuration and export objects prior to starting NFS servers.