Chapter 7. Deploying the Shared File Systems service with CephFS-NFS

With the Shared File Systems service (manila) with Ceph File System (CephFS) through NFS, you can use the same Red Hat Ceph Storage cluster that you use for block and object storage to provide file shares through the NFS protocol. For more information, see Configuring the Shared File Systems service (manila) in the Storage Guide.

Important

The RHOSP Shared File Systems service with CephFS through NFS for RHOSP 17.0 and later is supported for use with Red Hat Ceph Storage version 5.x or later. For more information about how to determine the version of Ceph Storage installed on your system, see Red Hat Ceph Storage releases and corresponding Ceph package versions.

CephFS is the highly scalable, open-source distributed file system component of Red Hat Ceph Storage, a unified distributed storage platform. Ceph Storage implements object, block, and file storage using Reliable Autonomic Distributed Object Store (RADOS). CephFS, which is POSIX compatible, provides file access to a Ceph storage cluster.

The Shared File Systems service enables users to create shares in CephFS and access them with NFS 4.1 through NFS-Ganesha. NFS-Ganesha controls access to the shares and exports them to clients through the NFS 4.1 protocol.

The Shared File Systems service manages the life cycle of these shares from within RHOSP. When cloud administrators configure the service to use CephFS through NFS, these file shares come from the CephFS cluster, but are created and accessed as familiar NFS shares.

For more information about the Shared File Systems service, see Configuring the Shared File Systems service (manila) in the Storage Guide.

7.1. CephFS with NFS driver

The CephFS through NFS back end in the Shared File Systems service (manila) is composed of Ceph metadata servers (MDS), the CephFS through NFS gateway (NFS-Ganesha), and the Ceph cluster service components. The Shared File Systems service CephFS NFS driver uses NFS-Ganesha gateway to provide NFSv4 protocol access to CephFS shares. The Ceph MDS service maps the directories and file names of the file system to objects that are stored in RADOS clusters. NFS gateways can serve NFS file shares with different storage back ends, such as Ceph. The NFS-Ganesha service runs on the Controller nodes with the Ceph services.

Instances are booted with at least two NICs: one NIC connects to the project router and the second NIC connects to the StorageNFS network, which connects directly to the NFS-Ganesha gateway. The instance mounts shares by using the NFS protocol. CephFS shares that are hosted on Ceph OSD nodes are provided through the NFS gateway.

NFS-Ganesha improves security by preventing user instances from directly accessing the MDS and other Ceph services. Instances do not have direct access to the Ceph daemons.

cephfs nfs topology nfs driver

7.2. Ceph services and client access

In addition to the monitor, OSD, Rados Gateway (RGW), and manager services deployed when Ceph provides object and block storage, a Ceph metadata service (MDS) is required for CephFS and an NFS-Ganesha service is required as a gateway to native CephFS using the NFS protocol. For user-facing object storage, an RGW service is also deployed. The gateway runs the CephFS client to access the Ceph public network and is under administrative rather than end-user control.

NFS-Ganesha runs in its own container that interfaces both to the Ceph public network and to a new isolated network, StorageNFS. The composable network feature of Red Hat OpenStack Platform (RHOSP) director deploys this network and connects it to the Controller nodes. As the cloud administrator, you can configure the network as a Networking (neutron) provider network.

NFS-Ganesha accesses CephFS over the Ceph public network and binds its NFS service using an address on the StorageNFS network.

To access NFS shares, provision user VMs, Compute (nova) instances, with an additional NIC that connects to the Storage NFS network. Export locations for CephFS shares appear as standard NFS IP:<path> tuples that use the NFS-Ganesha server VIP on the StorageNFS network. The network uses the IP address of the user VM to perform access control on the NFS shares.

Networking (neutron) security groups prevent the user VM that belongs to project 1 from accessing a user VM that belongs to project 2 over the StorageNFS network. Projects share the same CephFS file system but project data path separation is enforced because user VMs can access files only under export trees: /path/to/share1/…, /path/to/share2/….

7.3. Shared File Systems service with CephFS through NFS fault tolerance

When Red Hat OpenStack Platform (RHOSP) director starts the Ceph service daemons, they manage their own high availability (HA) state and, in general, there are multiple instances of these daemons running. By contrast, in this release, only one instance of NFS-Ganesha can serve file shares at a time.

To avoid a single point of failure in the data path for CephFS through NFS shares, NFS-Ganesha runs on a RHOSP Controller node in an active-passive configuration managed by a Pacemaker-Corosync cluster. NFS-Ganesha acts across the Controller nodes as a virtual service with a virtual service IP address.

If a Controller node fails or the service on a particular Controller node fails and cannot be recovered on that node, Pacemaker-Corosync starts a new NFS-Ganesha instance on a different Controller node using the same virtual IP address. Existing client mounts are preserved because they use the virtual IP address for the export location of shares.

Using default NFS mount-option settings and NFS 4.1 or later, after a failure, TCP connections are reset and clients reconnect. I/O operations temporarily stop responding during failover, but they do not fail. Application I/O also stops responding but resumes after failover completes.

New connections, new lock-state, and so on are refused until after a grace period of up to 90 seconds during which time the server waits for clients to reclaim their locks. NFS-Ganesha keeps a list of the clients and exits the grace period earlier if all clients reclaim their locks.

Note

The default value of the grace period is 90 seconds. To change this value, edit the NFSv4 Grace_Period configuration option.

7.4. CephFS through NFS-Ganesha installation

A typical Ceph file system (CephFS) through NFS installation in a Red Hat OpenStack Platform (RHOSP) environment includes the following configurations:

  • OpenStack Controller nodes running containerized Ceph metadata server (MDS), Ceph monitor (MON), manila, and NFS-Ganesha services. Some of these services can coexist on the same node or can have one or more dedicated nodes.
  • Ceph storage cluster with containerized object storage daemons (OSDs) running on Ceph storage nodes.
  • An isolated StorageNFS network that provides access from projects to the NFS-Ganesha services for NFS share provisioning.
Important

The Shared File Systems service (manila) with CephFS through NFS fully supports serving shares to Red Hat OpenShift Container Platform through Manila CSI. This solution is not intended for large scale deployments. For important recommendations, see https://access.redhat.com/articles/6667651.

The Shared File Systems service (manila) provides APIs that allow the projects to request file system shares, which are fulfilled by driver modules. The driver for Red Hat CephFS, manila.share.drivers.cephfs.driver.CephFSDriver, means that you can use the Shared File Systems service as a CephFS back end. RHOSP director configures the driver to deploy the NFS-Ganesha gateway so that the CephFS shares are presented through the NFS 4.1 protocol.

Using RHOSP director to deploy the Shared File Systems service with a CephFS back end on the overcloud automatically creates the required storage network defined in the heat template. For more information about network planning, see Overcloud networks in the Director Installation and Usage guide.

Although you can manually configure the Shared File Systems service by editing its node /etc/manila/manila.conf file, RHOSP director can override any settings in future overcloud updates. The recommended method for configuring a Shared File Systems back end is through director. Use RHOSP director to create an extra StorageNFS network for storage traffic.

Note

Adding CephFS through NFS to an externally deployed Ceph cluster, which was not configured by director, is supported. Currently, you can only define one CephFS back end in director. For more information, see Integrating an overcloud with Ceph Storage in Integrating an Overcloud with an Existing Red Hat Ceph Storage Cluster.

7.5. CephFS through NFS-Ganesha installation requirements

CephFS through NFS has been fully supported since Red Hat OpenStack Platform version (RHOSP) 13. The RHOSP Shared File Systems service with CephFS through NFS for RHOSP 17.0 and later is supported for use with Red Hat Ceph Storage version 5.2 or later. For more information about how to determine the version of Ceph Storage installed on your system, see Red Hat Ceph Storage releases and corresponding Ceph package versions.

Prerequisites

  • You install the Shared File Systems service on Controller nodes, as is the default behavior.
  • You install the NFS-Ganesha gateway service on the Pacemaker cluster of the Controller node.
  • You configure only a single instance of a CephFS back end to use the Shared File Systems service. You can use other non-CephFS back ends with the single CephFS back end.

7.6. File shares

File shares are handled differently between the OpenStack Shared File Systems service (manila), Ceph File System (CephFS), and Ceph through NFS.

The Shared File Systems service provides shares, where a share is an individual file system namespace and a unit of storage with a defined size. Shared file system storage inherently allows multiple clients to connect, read, and write data to any given share, but you must give each client access to the share through the Shared File Systems service access control APIs before they can connect.

With CephFS, a share is considered a directory with a defined quota and a layout that points to a particular storage pool or namespace. CephFS quotas limit the size of a directory to the size share that the Shared File Systems service creates. Access to CephFS through NFS shares is provided by specifying the IP address of the client.

With CephFS through NFS, file shares are provisioned and accessed through the NFS protocol. The NFS protocol also handles security.

7.7. Generating the custom roles file

For security, isolate NFS traffic to a separate network when using CephFS through NFS so that the Ceph NFS server is accessible only through the isolated network. Deployers can restrict the isolated network to a select group of projects in the cloud. Red Hat OpenStack (RHOSP) director ships with support to deploy a dedicated StorageNFS network. To configure and use the StorageNFS network, you require a custom Controller role.

Important

It is possible to omit the creation of an isolated network for NFS traffic. However, if you omit the StorageNFS network in a production deployment that has untrusted clients, director can connect the Ceph NFS server on any shared non-isolated network, such as an external network. Shared non-isolated networks are usually routable to all user private networks in the cloud. When the NFS server is on a non-isolated network, you cannot control access to Shared File Systems service (manila) shares by applying client IP access rules. Users must allow access to their shares by using the generic 0.0.0.0/0 IP. Because of the generic IP, anyone who discovers the export path can mount the shares.

You configure the isolated StorageNFS network by using the ControllerStorageNFS custom role file. This role file is similar to the default Controller.yaml role file, with the addition of the StorageNFS network.

[stack@undercloud ~]$ cd /usr/share/openstack-tripleo-heat-templates/roles
[stack@undercloud roles]$ diff Controller.yaml ControllerStorageNfs.yaml
16a17
> 	- StorageNFS
50a45

The openstack overcloud roles generate command creates a custom roles_data.yaml file, including the services specified after -o. In the following example, the roles_data.yaml file created has the services for ControllerStorageNfs, Compute, and CephStorage.

For more information about the openstack overcloud roles generate command, see Composable services and custom roles in Director Installation and Usage.

Note

If you have an existing roles_data.yaml file, modify it to add ControllerStorageNfs, Compute, and CephStorage services to the configuration file.

Procedure

  1. Log in to the undercloud host as the stack user.
  2. Source the stackrc undercloud credentials file:

    $ source ~/stackrc
  3. Create the roles_data.yaml file:

    [stack@undercloud ~]$ openstack overcloud roles generate \
    --roles-path /usr/share/openstack-tripleo-heat-templates/roles \
    -o /home/stack/roles_data.yaml ControllerStorageNfs Compute CephStorage

7.8. Deploying the updated environment

When you are ready to deploy your environment, use the openstack overcloud deploy command with the custom environments and roles required to run CephFS with NFS-Ganesha.

The overcloud deploy command has the following options in addition to other required options.

ActionOptionAdditional information

Add the extra StorageNFS network with network_data_ganesha.yaml

-n /usr/share/openstack-tripleo-heat-templates/network_data_ganesha.yaml

You can omit this option if you do not want to isolate NFS traffic to a separate network.

Add the custom roles defined in the roles_data.yaml file from the previous section

-r /home/stack/roles_data.yaml

You can omit this option if you do not want to isolate NFS traffic to a separate network.

Deploy the Ceph daemons.

-e /usr/share/openstack-tripleo-heat-templates/environments/cephadm/cephadm.yaml

Initiating overcloud deployment in Deploying Red Hat Ceph Storage and Red Hat OpenStack Platform together with director

Deploy the Ceph metadata server with ceph-mds.yaml

-e /usr/share/openstack-tripleo-heat-templates/environments/cephadm/ceph-mds.yaml

Initiating overcloud deployment in Deploying Red Hat Ceph Storage and Red Hat OpenStack Platform together with director

Deploy the Shared File Systems service (manila) with the CephFS-NFS back end. Configure NFS-Ganesha with director.

-e /usr/share/openstack-tripleo-heat-templates/environments/manila-cephfsganesha-config.yaml

The manila-cephfsganesha-config.yaml environment file

The following example shows an openstack overcloud deploy command with options to deploy CephFS through NFS-Ganesha, Ceph Storage cluster, Ceph MDS, and the isolated StorageNFS network:

[stack@undercloud ~]$ openstack overcloud deploy \
--templates /usr/share/openstack-tripleo-heat-templates  \
-n /usr/share/openstack-tripleo-heat-templates/network_data_ganesha.yaml \
-r /home/stack/roles_data.yaml \
-e /home/stack/containers-default-parameters.yaml   \
-e /usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml   \
-e /home/stack/network-environment.yaml  \
-e /usr/share/openstack-tripleo-heat-templates/environments/cephadm/cephadm.yaml  \
-e /usr/share/openstack-tripleo-heat-templates/environments/cephadm/ceph-mds.yaml  \
-e /usr/share/openstack-tripleo-heat-templates/environments/manila-cephfsganesha-config.yaml

For more information about the openstack overcloud deploy command, see Provisioning and deploying your overcloud in Director Installation and Usage.

7.8.1. The StorageNFS and network_data_ganesha.yaml file

Use composable networks to define custom networks and assign them to any role. Instead of using the standard network_data.yaml file, you can configure the StorageNFS composable network with the network_data_ganesha.yaml file. Both of these roles are available in the /usr/share/openstack-tripleo-heat-templates directory.

IMPORTANT
If you do not define the Storage NFS network, director defaults to the external network. Although the external network can be useful in test and prototype environments, security on the external network is not sufficient for production environments. For example, if you expose the NFS service on the external network, a denial of service (DoS) attack can disrupt controller API access to all cloud users, not only consumers of NFS shares. By contrast, when you deploy the NFS service on a dedicated Storage NFS network, potential DoS attacks can target only NFS shares in the cloud. In addition to potential security risks, when you deploy the NFS service on an external network, additional routing configurations are required for precise access control to shares. On the Storage NFS network, however, you can use the client IP address on the network to achieve precise access control.

The network_data_ganesha.yaml file contains an additional section that defines the isolated StorageNFS network. Although the default settings work for most installations, you must edit the YAML file to add your network settings, including the VLAN ID, subnet, and other settings.

name: StorageNFS
enabled: true
vip: true
name_lower: storage_nfs
vlan: 70
ip_subnet: '172.17.0.0/20'
allocation_pools: [{'start': '172.17.0.4', 'end': '172.17.0.250'}]
ipv6_subnet: 'fd00:fd00:fd00:7000::/64'
ipv6_allocation_pools: [{'start': 'fd00:fd00:fd00:7000::4', 'end': 'fd00:fd00:fd00:7000::fffe'}]

For more information about composable networks, see Composable networks in Director Installation and Usage.

7.8.2. CephFS-NFS back-end environment file

The environment file for defining a CephFS-NFS back end, manila-cephfsganesha-config.yaml, is located in the following path of an undercloud node: /usr/share/openstack-tripleo-heat-templates/environments/manila-cephfsganesha-config.yaml.

The manila-cephfsganesha-config.yaml environment file contains settings relevant to the deployment of the Shared File Systems service (manila). The back-end default settings work for most environments. The following example shows the default values that director uses during deployment of the Shared File Systems service:

[stack@undercloud ~]$ cat /usr/share/openstack-tripleo-heat-templates/environments/manila-cephfsganesha-config.yaml
# A Heat environment file which can be used to enable a
# a Manila CephFS-NFS driver backend.
resource_registry:
  OS::TripleO::Services::ManilaApi: ../deployment/manila/manila-api-container-puppet.yaml
  OS::TripleO::Services::ManilaScheduler: ../deployment/manila/manila-scheduler-container-puppet.yaml
  # Only manila-share is pacemaker managed:
  OS::TripleO::Services::ManilaShare: ../deployment/manila/manila-share-pacemaker-puppet.yaml
  OS::TripleO::Services::ManilaBackendCephFs: ../deployment/manila/manila-backend-cephfs.yaml
  # ceph-nfs (ganesha) service is installed and configured by Director
  # but it's still managed by pacemaker
  OS::TripleO::Services::CephNfs: ../deployment/cephadm/ceph-nfs.yaml

parameter_defaults:
  ManilaCephFSBackendName: cephfs 1
  ManilaCephFSDriverHandlesShareServers: false 2
  ManilaCephFSCephFSAuthId: 'manila' 3
  # manila cephfs driver supports either native cephfs backend - 'CEPHFS'
  # (users mount shares directly from ceph cluster), or nfs-ganesha backend -
  # 'NFS' (users mount shares through nfs-ganesha server)
  ManilaCephFSCephFSProtocolHelperType: 'NFS'

The parameter_defaults header signifies the start of the configuration. To override default values set in resource_registry, copy this manila-cephfsganesha-config.yaml environment file to your local environment file directory, /home/stack/templates/, and edit the parameter settings as required by your environment. This includes values set by OS::Tripleo::Services::ManilaBackendCephFs, which sets defaults for a CephFS back end.

1
ManilaCephFSBackendName sets the name of the manila configuration of your CephFS back end. In this case, the default back-end name is cephfs.
2
ManilaCephFSDriverHandlesShareServers controls the lifecycle of the share server. When set to false, the driver does not handle the lifecycle. This is the only supported option.
3
ManilaCephFSCephFSAuthId defines the Ceph auth ID that director creates for the manila service to access the Ceph cluster.

For more information about environment files, see Environment files in Director Installation and Usage.