Release Notes #

ceph-salt and cephadm.  SUSE Enterprise Storage 7 introduces a new deployment stack that is built on two tools: ceph-salt and cephadm. ceph-salt has been introduced to facilitate Day 1 Ceph cluster deployment. Like the DeepSea tool that it replaces, it uses Salt to prepare the cluster nodes and bootstrap a Ceph cluster on them. Unlike DeepSea, this tool only can only deploy a single Ceph Monitor daemon and a single Ceph Manager daemon on one node (this process is called bootstrapping). The rest of the deployment (additional Ceph Monitors and Managers, OSDs, Gateways, etc.) is handled by cephadm. cephadm has been introduced to allow Day 2 Ceph daemon deployment and management via containers into the orchestration layer. For more information, see https://documentation.suse.com/ses/7/html/ses-all/deploy-cephadm.html#deploy-cephadm-day2.

Ceph release.  SUSE Enterprise Storage 7 is based on Ceph Octopus v15.2.

Technology Previews, see Section 3, “Technology Previews”

Sound, graphics, fonts and artwork

Packages that require an additional customer contract

Packages with names ending in -devel (containing header files and similar developer resources) will only be supported together with their main packages.

Allows managing ceph-core (mgr, mon, osd), gateway (nfs-ganesha, mds, rgw), and the monitoring stack in a declarative way.

Feature parity with DeepSea.

The PG autoscaler feature introduced in Nautilus is enabled for new pools by default, allowing new clusters to autotune pg num without any user intervention. The default values for new pools and RGW/CephFS metadata pools have also been adjusted to perform well for most users.

Health alerts are now raised for recent Ceph daemons crashes.

A simple alerts module has been introduced to send email health alerts for clusters deployed without the benefit of an existing external monitoring infrastructure.

Automatic configuration of Prometheus, Alertmanager and Grafana to talk to each other and the Ceph Dashboard.

Per RBD graphs can now be displayed in Ceph Dashboard.

Ceph exporter (Prometheus manager module) enhancements: Cache overhauled and performance logs added to identify bottlenecks.

Customization of monitoring component configuration files is implemented through Jinja2 templates.

RADOS objects can now be brought in sync during recovery by copying only the modified portion of the object, reducing tail latencies during recovery.

RADOS snapshot trimming metadata is now managed in a more efficient and scalable fashion.

Now when noscrub and/or no deep-scrub flags are set globally or per pool, scheduled scrubs of the type disabled will be aborted. All user initiated scrubs are NOT interrupted.

BlueStore has received several improvements and performance updates, including improved accounting for omap (key/value) object data by pool, improved cache memory management, and a reduced allocation unit size for SSD devices. (Note that by default, the first time each OSD starts after upgrading to octopus it will trigger a conversion that may take from a few minutes to a few hours, depending on the amount of stored omap data.)

Ceph will allow recovery below min_size for erasure-coded pools, wherever possible.

NFS v4.1 protocol and newer is supported; NFS v3 is not supported

RADOS grace is now a supported recovery back-end

Now supports per-service configuration via a RADOS common configuration object

NFS Ganesha for FSAL Ceph now responds to cache pressure requests from libcephfs

The name of the RBD pool object that is used to store RBD trash purge schedule is changed from rbd_trash_trash_purge_schedule to rbd_trash_purge_schedule.

If you are already using RBD trash purge schedule functionality and have per-pool or per-namespace schedules configured, before the upgrade: Copy the rbd_trash_trash_purge_schedule object to rbd_trash_purge_schedule. Remove rbd_trash_purge_schedule using the following commands in every RBD pool and namespace where a trash purge schedule was previously configured:

> rados -p <pool-name> [-N namespace] cp \
  rbd_trash_trash_purge_schedule rbd_trash_purge_schedule

> rados -p <pool-name> [-N namespace] rm rbd_trash_trash_purge_schedule

Alternatively, restore the schedule after the upgrade in another way.

RBD mirroring now supports a new snapshot-based mode that no longer requires the journaling feature and its related impacts in exchange for the loss of point-in-time consistency (it remains crash-consistent).

RBD clone operations now preserve the sparseness of the underlying RBD image.

The RBD trash feature has been improved to (optionally) automatically move old parent images to the trash when their children are all deleted or flattened. It can now be configured to automatically purge on a defined schedule.

RBD images can be online re-sparsified to reduce the usage of zeroed extents.

The rbd-nbd tool has been improved to use more modern kernel interfaces. Caching has been improved to be more efficient and performant. The rbd-mirror daemon now automatically adjusts its per-image memory usage based upon its memory target.

Now supports the S3 Object lock feature, providing a write-once-read-many (WORM)-like functionality.

S3 API support for key value pairs on buckets similar to objects.

Automatic static subtree partitioning policies may now be configured using the new distributed and random ephemeral pinning extended attributes on directories.

MDS daemons can now be assigned to manage a particular file system via the new mds_join_fs.

MDS now aggressively asks idle clients to trim caps which improves stability when file system load changes.

Updated Samba and CTDB packages are carried on the SES media. These packages should be used for Samba Gateway deployments instead of base SLES packages.

The new version 4.13 Samba release provides significant performance improvements for encrypted SMB3 workloads.

CTDB now advertises clustered Samba presence in the Ceph Manager service map

New page layout and branding. The Dashboard UI now uses a layout with a vertical navigation bar to the left that can be hidden to free up some screen real estate. The branding has been updated to match SUSE's new corporate CI.

A new unified tasks and notifications bar that shows both ongoing activity and background tasks running in the cluster as well as past notifications. These can be removed individually or all at once.

Many pages now support multi-row selection, to perform bulk actions on some or all elements at once.

Tables provide custom filters that enable you to further drill down into the data shown.

User accounts can now be disabled temporarily or permanently without the need to delete them.

It is now also possible to force users to change their initial password at the first login. Users can also change their passwords via the Dashboard without administrator intervention at any time.

The Dashboard can also enforce a variety of password complexity rules if required, or let passwords expire after a configurable amount of time.

Most of these password features are disabled by default. They can be enabled and configured individually to help adhering to any local password security policies that may be in force.

It is now possible to clone existing user roles to save time when creating new ones that are similar to already existing roles.

cephadm/orchestrator integration: Display and filter information about the cluster and its inventory from a per-host, a device, or service perspective.

Deploy new OSDs via cephadm based on customizable filter rules.

Display all hosts known to the cluster and all their devices and services.

Obtain individual health status and SMART data.

Host inventory shows all disks attached, as well as their type, size and other details.

The host page shows all services that have been deployed on the selected host, which container they are running in and their current status.

Clicking the button Identify will help finding the selected device in a data center, by making the disk enclosure LED blink for a customizable amount of time.

When creating a new pool, it is now possible to define CRUSH placement rules to specify device classes, so fast pools can be created on SSDs only, for example. This helps in creating different tiers of storage pools easily.

It is now possible to define the PG autoscaling mode on a per-pool basis, either by choosing the cluster's default, or selecting a different mode. For example, you could disable autoscaling or only emit a warning when the pool’s PG count is not ideal.

Defining per-pool quotas for both the amount of data in total or the number of objects that can be stored in a pool.

Object Gateway management now supports some new RGW features like versioned buckets, multi-factor authentication and the selection of placement targets when creating buckets.

On the CephFS management page, it is now possible to disconnect or “evict” clients from the list of active sessions and to create snapshots of a CephFS subtree manually.

Support for managing CephFS quotas was added as well as a simple filesystem browser that allows users to traverse the file system’s directory structure.

The iSCSI management pages were also improved, giving a more detailed insight into active iSCSI gateways and initiators now. Moreover, some safeguards around deleting IQNs with active sessions were added.

The integration with the Prometheus alert manager was enhanced, now showing all configured alerts, not just the current active ones.

A dedicated page/workflow was added to enable the submission of telemetry data.