Managing Remote Mounts¶
This section describes how you can mount a NFS/CIFS location inside the Masking Engine and use it in a masking job.
In order to access the files shared over NFS/CIFS server from the Masking Engine, complete the following two steps:
- Create and connect a mount using Mount Filesystem API endpoint.
- Create a file connector with Filesystem Mount Point mode. Or, Upload a XML/Copybook file format using Filesystem Mount Point mode.
Mount Filesystem API¶
The Mount Filesystem APIs are used to perform normal CRUD operations(Create, Read, Update and Delete) along with three mount operations connect(mount), disconnect(unmount) and remount on a mount object.
To create a mount entry, information about the mount are passed. Some of them are required and some are optional.
- Require Information:
- mountName: The name of the mount. This name is used to refer this mount in the connector creation and file format upload UIs.
- hostAddress: The NFS/CIFS server address.
- mountPath: The remote path shared by the NFS/CIFS servers.
- type: The type of the server. CIFS, NFS3, or NFS4.
- Optional Information:
- options: The mount options.
- connectOnStartup: Whether this mount should be connected or not when the server starts.
When a server shuts down, all the mounts are disconnected.
The API supports passing many mount options. Not all of them are supported by a server. After a mount is connected, you might see the options field has many options that were not passed by you or some options have been eliminated that were passed by you. The options field shows effective options only. The applied options are gathered after a mount is connected.
The API also restricts usage of some mount options.
The following mount options are enforced and added to the list of options for all mounts:
- nosuid: The filesystem cannot contain set userid files.
- noexec: No executable script can be run from the mount.
- nodev: The filesystem cannot contain special devices.
options is an optional field, it is required for CIFS mounts to pass credentials. the following options are required for CIFS mounts:
- username: The username to connect to the CIFS server.
- password: The password of the user.
- domain: The domain of the user.
For NFSv3 mounts,
options are not required, therefore can be
For NFSv4 mounts, the following option is required:
- nfsvers: The NFS protocol version number. For example, "options": "nfsvers=4.0"
The version information is passed using vers option. The supported versions based on mount types are
|Mount Type||Supported Versions|
|CIFS||2.0, 2.1, 3.0|
|NFS4||4, 4.0, 4.1, 4.2|
Some mount options are generic which can be applied to all the mount types while some are mount specific options. In the case of remount operation, only generic options can be modified. The list of allowed generic options are:
async, atime, auto, context, defaults, defcontext, diratime, dirsync, fscontext, group, iversion, lazytime, loud, mand, _netdev, noatime, noauto, nodev, nodiratime, noexec, nofail, noiversion, nolazytime, nomand, norelatime, nostrictatime, nosuid, nouser, owner, relatime, _rnetdev, ro, rootcontext, rw, silent, strictatime, sync, and user.
The create endpoint is used to create a mount entry. It takes all the information about a mount as its input and creates a mount entry. It doesn't do any kind of validation about the mount's accessibility. The validation is done during the connect operation.
The read endpoints are used to retrieve information about a mount. There are two read endpoints. 1. get all: To get information about all mounts. 2. get: To get information about any particular mount identified by its id.
The update endpoint is used to modify any information of a mount. Update operation can be performed only on a disconnected mount.
The delete endpoint is used to delete a mount entry. A mount can be deleted only if it is not being used in any of the connectors.
Apart from normal CRUD operations, there are three special mount related operations exposed through the API.
The connect endpoint is used to mount a remote mount inside the masking engine. If the connect operation succeeds then, the options field is updated with the applied mount options.
The disconnect endpoint is used to unmount a remote mount from the Masking Engine.
The API supports the remount operation. This can be used to remount an active or to connect a disconnected mount and also to update some mount information. This can update mountName, connectOnStartup and generic options only. For other updates, use the normal update API.
Resolve Mount Consistency¶
A script runs in the background to keep the data in the mount_information table and mounts in sync. If for some reason, the data for a mount mounted inside the mount engine and data corresponding to that mount in mount_information table becomes inconsistent, the mount is unmounted. For example, if a mount is in a disconnected state in DB but it is mounted in the engine, then it will be unmounted.
A mount can be used at two places:
- File connectors
- File formats
While creating a connector, when any file connector option is selected, the UI shows a dropdown to select how a file will be accessed. There are three options:
- Filesystem Mount Point
On selecting the Filesystem Mount Point option, the mount name and a path inside the mount needs to be specified.
- Mount Name: This is a list of mount names created in the engine.
- Path Under Mount: A path relative to the path mounted. By default, it is at the root of the remote Mount path.
- Remote Path: The complete remote path. On selecting a mount name and typing a path in the above input box, this gets updated.
A connector can be created even if a mount is in a disconnected state but it should be in an active state when a ruleset is being created or when a job is run.
The XML and Copybook file formats can be uploaded from a remote location. To upload a file format from an NFS/CIFS location, select the Filesystem Mount Point option.
A mount can be synced from a source engine to a target engine using Sync APIs. Syncing a file connector using a mount also syncs the related mounts. The following mount information fields are synced:
In case of CIFS mounts, the password is not synced. In order to set the password in the target engine, update the mount's options and ensure to include the password in the options.
Recommended Mount Server Configuration¶
The NFS and CIFS servers should be configured in such a way that the files are readable and writeable by the Masking Engine.
The user-provided to connect to the mount should have read and write permission on the mount.
- The Masking Engine's server IP should have read and write permission on the mount.
- For NFS, the access to a file is controlled based on the UID and GID. In order to give read & write permission to the Masking Engine on the share path, the path should be shared with the following options:
<mount path> <masking engine ip>(rw,all_squash,anonuid=<uid>,anongid=<gid>) # uid and gid is of the owner of the shared path on the server