From bf3a83b46c62cdb1bb1ff5d956dd5a207d4a4044 Mon Sep 17 00:00:00 2001 From: Carlos Eduardo Date: Wed, 18 Sep 2024 16:46:38 -0300 Subject: [PATCH] Add CephFS Manage/unmanage documentation - Updates the drivers support matrix - Updates the CephFS drivers documentation and highlight how the process works and what are the known limitations. Change-Id: I2f937358e93271fc5ad63e76cc4a9c5aa0ba3017 (cherry picked from commit 4fe2b014e52d095d333e4cdb7ffb3269974bde43) --- ...hare_back_ends_feature_support_mapping.rst | 2 +- .../drivers/cephfs_driver.rst | 24 +++++++++++++++++++ 2 files changed, 25 insertions(+), 1 deletion(-) diff --git a/doc/source/admin/share_back_ends_feature_support_mapping.rst b/doc/source/admin/share_back_ends_feature_support_mapping.rst index b4627e72fe..b23ed1d848 100644 --- a/doc/source/admin/share_back_ends_feature_support_mapping.rst +++ b/doc/source/admin/share_back_ends_feature_support_mapping.rst @@ -89,7 +89,7 @@ Mapping of share drivers and share features support +----------------------------------------+-----------------------+-----------------------+--------------------------+--------------------------+------------------------+-----------------------------------+--------------------------+--------------------+--------------------+ | Oracle ZFSSA | K | N | M | M | K | K | \- | \- | \- | +----------------------------------------+-----------------------+-----------------------+--------------------------+--------------------------+------------------------+-----------------------------------+--------------------------+--------------------+--------------------+ -| CephFS | M | \- | M | M | M | W | \- | \- | \- | +| CephFS | M | D (2024.2) | M | M | M | W | D (2024.2) | \- | \- | +----------------------------------------+-----------------------+-----------------------+--------------------------+--------------------------+------------------------+-----------------------------------+--------------------------+--------------------+--------------------+ | Tegile | M | \- | M | M | M | M | \- | \- | \- | +----------------------------------------+-----------------------+-----------------------+--------------------------+--------------------------+------------------------+-----------------------------------+--------------------------+--------------------+--------------------+ diff --git a/doc/source/configuration/shared-file-systems/drivers/cephfs_driver.rst b/doc/source/configuration/shared-file-systems/drivers/cephfs_driver.rst index e7975c5601..648d42388c 100644 --- a/doc/source/configuration/shared-file-systems/drivers/cephfs_driver.rst +++ b/doc/source/configuration/shared-file-systems/drivers/cephfs_driver.rst @@ -49,6 +49,8 @@ The following operations are supported with CephFS backend: * ``read-only`` and ``read-write`` access levels are supported. - Extend/shrink share +- Manage/unmanage shares +- Manage/unmanage share snapshots - Create, delete, update and list snapshot - Create, delete, update and list share groups - Delete and list share group snapshots @@ -511,6 +513,28 @@ server and the path to be mounted. It is of the form, ``{NFS-Ganesha server address}:{path to be mounted}`` +Managing existing shares and snapshots +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Starting from the 2024.2 (Dalmatian) release, it is possible to bring both +CephFS Native and NFS shares (subvolumes) that already exist in the Ceph +Storage under Manila's management. The workflow will be slightly different when +compared to the original approach: + +- The :ref:`share adoption process` should be used, but the name + of the subvolume must be supplied as the ``export path`` parameter, so that + the driver can appropriately locate the subvolume and manage it. +- The subvolume's ``subvolume_mode`` will not be overwritten by manila. +- The subvolumes will not be renamed in the Ceph Storage, and at the end of the + manage operation, their names will be preserved. +- In case the subvolume has its current allocated size (``bytes_quota``) set + as infinite, please make sure specify a new ``size`` within the + ``driver_options`` parameter. The driver will attempt to resize the share + using the provided size and will fail if it is not enough. +- While managing CephFS Snapshots, the CephFS driver will consider the size of + the parent share and account it as the snapshot's quota. + + Allowing access to shares ~~~~~~~~~~~~~~~~~~~~~~~~~