d/numam-spdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

lib/blob: document use_extent_table

Browse Source 
Patch adds documentation and CHANGELOG update
for newly added Extent Table/Page path.

Change-Id: I86f6c5680084a92d50bd9ca39b68d68a9908ecf8
Signed-off-by: Tomasz Zawadzki <tomasz.zawadzki@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/483247
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Darek Stojaczyk <dariusz.stojaczyk@intel.com>
Community-CI: SPDK CI Jenkins <sys_sgci@intel.com>
This commit is contained in:
Tomasz Zawadzki 2020-01-29 05:21:05 -05:00
parent d0f4a51fdc
commit 353252b1b4
2 changed files with 25 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
CHANGELOG.md
View File

@ -24,6 +24,13 @@ Added boolean return value for function spdk_fs_set_cache_size to indicate its o
Added `blobfs_set_cache_size` RPC method to set cache size for blobstore filesystem.
### blobstore
Added new `use_extent_table` option to `spdk_blob_opts` for creating blobs with Extent Table descriptor.
Using this metadata format, dramatically decreases number of writes required to persist each cluster allocation
for thin provisioned blobs. Extent Table descriptor is enabled by default.
See the [Blobstore Programmer's Guide](https://spdk.io/doc/blob.html#blob_pg_cluster_layout) for more details.
### dpdk
Updated DPDK submodule to DPDK 19.11.

18
doc/blob.md
View File

@ -318,6 +318,24 @@ form a linked list. The first page in the list will be written in place on updat
be written to fresh locations. This requires the backing device to support an atomic write size greater than
or equal to the page size to guarantee that the operation is atomic. See the section on atomicity for details.
### Blob cluster layout {#blob_pg_cluster_layout}
Each blob is an ordered list of clusters, where starting LBA of a cluster is called extent. A blob can be
thin provisioned, resulting in no extent for some of the clusters. When first write operation occurs
to the unallocated cluster - new extent is chosen. This information is stored in RAM and on-disk.
There are two extent representations on-disk, dependent on `use_extent_table` (default:true) opts used
when creating a blob.
* **use_extent_table=true**: EXTENT_PAGE descriptor is not part of linked list of pages. It contains extents
that are not run-length encoded. Each extent page is referenced by EXTENT_TABLE descriptor, which is serialized
as part of linked list of pages. Extent table is run-length encoding all unallocated extent pages.
Every new cluster allocation updates a single extent page, in case when extent page was previously allocated.
Otherwise additionally incurs serializing whole linked list of pages for the blob.
* **use_extent_table=false**: EXTENT_RLE descriptor is serialized as part of linked list of pages.
Extents pointing to contiguous LBA are run-length encoded, including unallocated extents represented by 0.
Every new cluster allocation incurs serializing whole linked list of pages for the blob.
### Sequences and Batches
Internally Blobstore uses the concepts of sequences and batches to submit IO to the underlying device in either