> block number (e.g., the location on disk). It's a cache; lookups are > fast, and is an in-memory lookup. Well, it's a little more than a > cache, it also stores some information for delayed allocation buffered > writes. For sequential access, it will traverse almost the same path of the tree. How deep the extent status tree be in general? If the tree is much deeper, the sequential accesses would have many repeated nodes traversal on the tree for the lookup. Have you observed significant bottleneck on "ext4_es_lookup_extent"? Can it be removed by caching the parent node? -- Probir Roy
