ceph/doc/rados/configuration/journal-ref.rst

   1 ==========================
   2  Journal Config Reference
   3 ==========================
   4 .. warning:: Filestore has been deprecated in the Reef release and is no longer supported.
   5 .. index:: journal; journal configuration
   6
   7 Filestore OSDs use a journal for two reasons: speed and consistency.  Note
   8 that since Luminous, the BlueStore OSD back end has been preferred and default.
   9 This information is provided for pre-existing OSDs and for rare situations where
  10 Filestore is preferred for new deployments.
  11
  12 - **Speed:** The journal enables the Ceph OSD Daemon to commit small writes
  13   quickly. Ceph writes small, random i/o to the journal sequentially, which
  14   tends to speed up bursty workloads by allowing the backing file system more
  15   time to coalesce writes. The Ceph OSD Daemon's journal, however, can lead
  16   to spiky performance with short spurts of high-speed writes followed by
  17   periods without any write progress as the file system catches up to the
  18   journal.
  19
  20 - **Consistency:** Ceph OSD Daemons require a file system interface that
  21   guarantees atomic compound operations. Ceph OSD Daemons write a description
  22   of the operation to the journal and apply the operation to the file system.
  23   This enables atomic updates to an object (for example, placement group
  24   metadata). Every few seconds--between ``filestore max sync interval`` and
  25   ``filestore min sync interval``--the Ceph OSD Daemon stops writes and
  26   synchronizes the journal with the file system, allowing Ceph OSD Daemons to
  27   trim operations from the journal and reuse the space. On failure, Ceph
  28   OSD Daemons replay the journal starting after the last synchronization
  29   operation.
  30
  31 Ceph OSD Daemons recognize the following journal settings:
  32
  33 .. confval:: journal_dio
  34 .. confval:: journal_aio
  35 .. confval:: journal_block_align
  36 .. confval:: journal_max_write_bytes
  37 .. confval:: journal_max_write_entries
  38 .. confval:: journal_align_min_size
  39 .. confval:: journal_zero_on_create