配置选项¶

通用 IO 选项¶

rbd compression hint

描述: Hint to send to the OSDs on write operations. If set to compressible and the OSD bluestore compression mode setting is passive, the OSD will attempt to compress the data. If set to incompressible and the OSD compression setting is aggressive, the OSD will not attempt to compress the data.
类型: Enum
是否必需: No
默认值: none
Values: none, compressible, incompressible

rbd read from replica policy

描述: policy for determining which OSD will receive read operations. If set to default, the primary OSD will always be used for read operations. If set to balance, read operations will be sent to a randomly selected OSD within the replica set. If set to localize, read operations will be sent to the closest OSD as determined by the CRUSH map. Note: this feature requires the cluster to be configured with a minimum compatible OSD release of Octopus.
类型: Enum
是否必需: No
默认值: default
Values: default, balance, localize

缓存选项¶

Ceph 块设备的用户空间实现（即 librbd ）不能利用 Linux 页缓存，所以它自己实现了内存缓存，名为“ RBD 缓存”。 RBD 缓存就像硬盘缓存一样工作，当 OS 发送了 barrier 或 flush 请求时，所有脏数据都会写入 OSD ，这意味着只要 VM 会正确地发送回写命令（即内核版本大于 2.6.32 ），使用回写缓存就像用正常的物理硬盘一样安全。此缓存用最近最少使用（ Least Recently Used, LRU ）算法，而且在回写模式下它能合并相邻请求以提高吞吐量。

The librbd cache is enabled by default and supports three different cache policies: write-around, write-back, and write-through. Writes return immediately under both the write-around and write-back policies, unless there are more than rbd cache max dirty unwritten bytes to the storage cluster. The write-around policy differs from the write-back policy in that it does not attempt to service read requests from the cache, unlike the write-back policy, and is therefore faster for high performance write workloads. Under the write-through policy, writes return only when the data is on disk on all replicas, but reads may come from the cache.

Prior to receiving a flush request, the cache behaves like a write-through cache to ensure safe operation for older operating systems that do not send flushes to ensure crash consistent behavior.

If the librbd cache is disabled, writes and reads go directly to the storage cluster, and writes return only when the data is on disk on all replicas.

Note

The cache is in memory on the client, and each RBD image has its own. Since the cache is local to the client, there’s no coherency if there are others accessing the image. Running GFS or OCFS on top of RBD will not work with caching enabled.

RBD 选项应该位于 ceph.conf 配置文件的 [client] 段下，可用选项有：

rbd cache

描述: 允许为 RADOS 块设备提供缓存。
类型: Boolean
是否必需: No
默认值: true

rbd cache policy

描述: Select the caching policy for librbd.
类型: Enum
是否必需: No
默认值: writearound
Values: writearound, writeback, writethrough

rbd cache writethrough until flush

描述: 开始进入 write-through 模式，并且在首个 flush 请求收到后切回 write-back 模式。启用它保守但安全，以防 rbd 之上的虚拟机内核太老、不能发送 flush ，像 2.6.32 之前的 virtio 驱动。
类型: Boolean
是否必需: No
默认值: true

rbd cache size

描述: RBD 缓存尺寸，字节。
类型: 64-bit Integer
是否必需: No
默认值: 32 MiB
适用策略: write-back and write-through

rbd cache max dirty

描述: 使缓存触发写回的 dirty 临界点，若为 0 ，直接使用写透缓存。
类型: 64-bit Integer
是否必需: No
约束条件: 必须小于 rbd cache size 。
默认值: 24 MiB
适用策略: write-around and write-back

rbd cache target dirty

描述: 缓存开始写回数据的目的地 dirty target ，不会阻塞到缓存的写动作。
类型: 64-bit Integer
是否必需: No
约束条件: 必须小于 rbd cache max dirty.
默认值: 16 MiB
适用策略: write-back

rbd cache max dirty age

描述: 写回开始前，脏数据在缓存中的暂存时间。
类型: Float
是否必需: No
默认值: 1.0
适用策略: write-back

预读选项¶

librbd 支持预读或预取功能，以此优化小块的顺序读。此功能通常应该由访客操作系统（是虚拟机）处理，但是引导加载程序还不能进行高效的读。如果缓存功能停用、或策略为 write-around ，预读就会自动被禁用。

rbd readahead trigger requests

描述: 触发预读的顺序读请求数量。
类型: Integer
是否必需: No
默认值: 10

rbd readahead max bytes

描述: 预读请求最大尺寸，零为禁用预读。
类型: 64-bit Integer
是否必需: No
默认值: 512 KiB

rbd readahead disable after bytes

描述: 从 RBD 映像读取这么多字节后，预读功能将被禁用，直到关闭。这样访客操作系统启动后就可以接管预读了，设为 0 时则仍开启预读。
类型: 64-bit Integer
是否必需: No
默认值: 50 MiB

映像功能¶

RBD supports advanced features which can be specified via the command line when creating images or the default features can be specified via Ceph config file via ‘rbd_default_features = <sum of feature numeric values>’ or ‘rbd_default_features = <comma-delimited list of CLI values>’

Layering

描述: Layering enables you to use cloning.
内置值: 1
CLI 值: layering
哪版加入: v0.52 (Bobtail)
KRBD 支持情况: since v3.10
默认值: yes

Striping v2

描述: Striping spreads data across multiple objects. Striping helps with parallelism for sequential read/write workloads.
内置值: 2
CLI 值: striping
哪版加入: v0.55 (Bobtail)
KRBD 支持情况: since v3.10 (default striping only, “fancy” striping added in v4.17)
默认值: yes

Exclusive locking

描述: When enabled, it requires a client to get a lock on an object before making a write. Exclusive lock should only be enabled when a single client is accessing an image at the same time.
内置值: 4
CLI 值: exclusive-lock
哪版加入: v0.92 (Hammer)
KRBD 支持情况: since v4.9
默认值: yes

Object map

描述: Object map support depends on exclusive lock support. Block devices are thin provisioned—meaning, they only store data that actually exists. Object map support helps track which objects actually exist (have data stored on a drive). Enabling object map support speeds up I/O operations for cloning; importing and exporting a sparsely populated image; and deleting.
内置值: 8
CLI 值: object-map
哪版加入: v0.93 (Hammer)
KRBD 支持情况: since v5.3
默认值: yes

Fast-diff

描述: Fast-diff support depends on object map support and exclusive lock support. It adds another property to the object map, which makes it much faster to generate diffs between snapshots of an image, and the actual data usage of a snapshot much faster.
内置值: 16
CLI 值: fast-diff
哪版加入: v9.0.1 (Infernalis)
KRBD 支持情况: since v5.3
默认值: yes

Deep-flatten

描述: Deep-flatten makes rbd flatten work on all the snapshots of an image, in addition to the image itself. Without it, snapshots of an image will still rely on the parent, so the parent will not be delete-able until the snapshots are deleted. Deep-flatten makes a parent independent of its clones, even if they have snapshots.
内置值: 32
CLI 值: deep-flatten
哪版加入: v9.0.2 (Infernalis)
KRBD 支持情况: since v5.1
默认值: yes

Journaling

描述: Journaling support depends on exclusive lock support. Journaling records all modifications to an image in the order they occur. RBD mirroring utilizes the journal to replicate a crash consistent image to a remote cluster.
内置值: 64
CLI 值: journaling
哪版加入: v10.0.1 (Jewel)
KRBD 支持情况: no
默认值: no

Data pool

描述: On erasure-coded pools, the image data block objects need to be stored on a separate pool from the image metadata.
内置值: 128
哪版加入: v11.1.0 (Kraken)
KRBD 支持情况: since v4.11
默认值: no

Operations

描述: Used to restrict older clients from performing certain maintenance operations against an image (e.g. clone, snap create).
内置值: 256
哪版加入: v13.0.2 (Mimic)
KRBD 支持情况: since v4.16

Migrating

描述: Used to restrict older clients from opening an image when it is in migration state.
内置值: 512
哪版加入: v14.0.1 (Nautilus)
KRBD 支持情况: no

Non-primary

描述: Used to restrict changes to non-primary images using snapshot-based mirroring.
内置值: 1024
哪版加入: v15.2.0 (Octopus)
KRBD 支持情况: no

QOS 选项¶

librbd supports limiting per image IO, controlled by the following settings.

rbd qos iops limit

描述: The desired limit of IO operations per second.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos bps limit

描述: The desired limit of IO bytes per second.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos read iops limit

描述: The desired limit of read operations per second.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos write iops limit

描述: The desired limit of write operations per second.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos read bps limit

描述: The desired limit of read bytes per second.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos write bps limit

描述: The desired limit of write bytes per second.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos iops burst

描述: The desired burst limit of IO operations.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos bps burst

描述: The desired burst limit of IO bytes.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos read iops burst

描述: The desired burst limit of read operations.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos write iops burst

描述: The desired burst limit of write operations.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos read bps burst

描述: The desired burst limit of read bytes.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos write bps burst

描述: The desired burst limit of write bytes.
类型: Unsigned Integer
是否必需: No
默认值: 0

rbd qos iops burst seconds

描述: The desired burst duration in seconds of IO operations.
类型: Unsigned Integer
是否必需: No
默认值: 1

rbd qos bps burst seconds

描述: The desired burst duration in seconds of IO bytes.
类型: Unsigned Integer
是否必需: No
默认值: 1

rbd qos read iops burst seconds

描述: The desired burst duration in seconds of read operations.
类型: Unsigned Integer
是否必需: No
默认值: 1

rbd qos write iops burst seconds

描述: The desired burst duration in seconds of write operations.
类型: Unsigned Integer
是否必需: No
默认值: 1

rbd qos read bps burst seconds

描述: The desired burst duration in seconds of read bytes.
类型: Unsigned Integer
是否必需: No
默认值: 1

rbd qos write bps burst seconds

描述: The desired burst duration in seconds of write bytes.
类型: Unsigned Integer
是否必需: No
默认值: 1

rbd qos schedule tick min

描述: The minimum schedule tick (in milliseconds) for QoS.
类型: Unsigned Integer
是否必需: No
默认值: 50