Example Analysis of ceph sparse read
This article mainly shows you the "sample analysis of ceph sparse read", which is easy to understand and well-organized. I hope it can help you solve your doubts. Let the editor lead you to study and learn the article "sample Analysis of ceph sparse read".
Before you explain ceph's sparse read, you need to explain a concept: the fiemap of the file system.
One of the features already supported by ext4 is Extent-based file storage, abbreviated as fiemap.
Many modern file systems use extent instead of block to manage disks. Extent can effectively reduce metadata overhead.
Besides, ceph's sparse read.
Ceph implements sparse read based on fiemap features. For example:
1. Client issues an object request for sparse-read.
2. Osd receives the request and passes it to filestore. It uses FS_IOC_FIEMAP ioctl to determine whether the underlying file system supports fiemap, and also determines whether the ceph configuration option turns on fiemap by default.
3. If both options are supported, filestore further obtains extent map information through this ioctl, that is, which extent holds real data.
4. Filestore uses extent map to analyze which parts of the object actually allocate space on disk and store data, and then read them out and return them to the client.
Note:
1. Ext4 supports fiemap by default, but ceph turns off the fiemap option by default, because the test found that the bug of a certain kernel/fs. So all client sparse read take the same path as normal read inside ceph.
2. Sparse read can improve the performance of big reading, but there is no performance improvement for small reading, such as reading less than one block size (4096B). Therefore, we should avoid doing sparse read for primary reading.
The above is all the content of this article "sample Analysis of ceph sparse read". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!
