Technique for accelerating file deletion by preloading indirect blocks
View Patent ↗A system and method for accelerating file deletion by preloading indirect blocks. When processing level 1 (L1) indirect blocks, the file server issues a read request for the next N indirect blocks to be loaded into a buffer cache. The file server processes the indirect blocks from the buffer cache without requiring multiple single block disk accesses. Once the N indirect blocks have been processed, a read request for the next N L1 indirect blocks is then sent. The technique permits parallel processing by loading indirect blocks to be processed while previously cached indirect blocks are being processed.
1. A method for accelerating file deletion of a file served by a storage system, comprising:
invoking a file deletion operation to delete a selected file;
reading, in response to the file deletion operation, a high-level indirect block of a buffer tree of the selected file to locate one or more low-level indirect blocks referenced by the high-level indirect block;
issuing read requests, in response to locating the one or more low-level indirect blocks, for the one or more low-level indirect blocks;
processing a first low-level indirect block of the one or more low-level indirect blocks in parallel with executing the read requests;
preloading into a buffer cache memory a number of one or more low-level indirect blocks that were read while the first low-level block is being processed, wherein the number, of the one or more low-level indirect blocks to be preloaded, chosen to ensure that the one or more low-level indirect blocks do not overflow the buffer cache memory; and
freeing data blocks referenced by pointers of the first low-level indirect block in response to the file deletion operation to delete the data blocks.
2. The method of claim 1 , further comprising:
processing, by the storage system, the low-level indirect blocks referenced by the high-level indirect block.
3. The method of claim 1 , further comprising:
preloading the one or more low-level indirect blocks referenced by the high-level indirect block into the buffer cache of the storage system before the one or more low-level indirect blocks are needed for processing.
4. The method of claim 1 , further comprising:
processing the first low-level indirect block by clearing a set of pointers in the first low level indirect block.
5. The method of claim 1 , further comprising:
representing each low-level indirect block as a 4 KB block.
6. A storage system, comprising:
an operating system having a file deletion operation;
in response to the file deletion operation, reading a high-level indirect block of a buffer tree of a selected file to locate one or more low-level indirect blocks referenced by the high-level indirect blocks;
read requests to be issued in response to locating the one or more low-level indirect blocks;
a first low-level indirect block of the one or more low-level indirect blocks to be processed in parallel to executing the read requests;
a number of one or more low-level indirect blocks that were read to be preloaded into a buffer cache memory while the first low level indirect block is being processed, wherein the number, of the one or more low-level indirect blocks to be preloaded, chosen to ensure that the one or more low-level indirect blocks do not overflow the buffer cache memory; and
data blocks referenced by pointers of the first low-level indirect block are freed in response to the file deletion operation to delete the data blocks.
7. The system of claim 6 , further comprising:
the one or more low-level indirect blocks referenced by the high-level indirect block are processed by the storage system.
8. The system of claim 6 , further comprising:
the first low-level indirect block is processed by clearing a set of pointers in the first low level indirect block.
9. The system of claim 6 , further comprising:
each low-level indirect block represented as a 4 KB block.
10. A computer readable media, comprising:
said computer readable media containing instructions for execution on a processor for a method of accelerated file deletion of a file served by a storage system, the method having;
invoking a file deletion operation to delete a selected file;
reading, in response to the file deletion operation, a high-level indirect block of a buffer tree of the selected file to locate one or more low-level indirect blocks referenced by the high-level indirect block;
issuing read requests, in response to locating the one or more low-level indirect blocks, for the one or more low-level indirect blocks;
processing a first low-level indirect block of the one or more low-level indirect blocks in parallel to executing-the read requests;
preloading into a buffer cache memory a number of one or more low-level indirect blocks that were read while the first low-level indirect block is being processed, wherein the number, of the one or more low-level indirect blocks to be preloaded, chosen to ensure that the one or more low-level indirect blocks do not overflow the buffer cache memory; and
freeing data blocks referenced by pointers of the first low-level indirect block in response to the file deletion operation to delete the data blocks.