I would appreciate it if Al (or anyone else
knowledgeable, but Al has been helping us greatly)
would review my recent changes to orangefs's
d_revalidate.
They have been in our tree for a few weeks now, but I
haven't heard anything about them.
commit 99109822f5cbe6d530eb55193b25aa5348f6134d
orangefs: Fix revalidate.
commit fe88adc3661ff9eb2a9777277f9c3abf5909449f
orangefs: Only compare attributes specified in orangefs_inode_getattr.
commit 933287da750edefbf0f449750fd67b4fc6c10013
orangefs: Implement inode_operations->permission().
commit 237f8282c04ba81926f4dfc33cd2ca20bb0c50e7
orangefs: Do not retrieve size from servers unless it it necessary.
are the ones I'm thinking of.
I have been working on a set of patches that will
utilize the inode and dentry caches to reduce load on
the servers. OrangeFS can return inode data with
readdir. I want to extend the protocol to get this into
the kernel so the inode can be filled during readdir.
Then when/if userspace calls stat while the cache is
hot, we won't have to do a trip to the server.
I believe this will significantly improve ls and
directory operations in general performance for
OrangeFS.
Thanks,
-- Martin
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
