On Mon, Nov 15, 2010 at 03:21:00PM +1100, Nick Piggin wrote: > On Mon, Nov 15, 2010 at 12:00:27PM +1100, Dave Chinner wrote: > > On Fri, Nov 12, 2010 at 12:24:21PM +1100, Nick Piggin wrote: > > > On Wed, Nov 10, 2010 at 9:05 AM, Nick Piggin <npiggin@xxxxxxxxx> wrote: > > > > On Tue, Nov 09, 2010 at 09:08:17AM -0800, Linus Torvalds wrote: > > > >> On Tue, Nov 9, 2010 at 8:21 AM, Eric Dumazet <eric.dumazet@xxxxxxxxx> wrote: > > > >> > > > > >> > You can see problems using this fancy thing : > > > >> > > > > >> > - Need to use slab ctor() to not overwrite some sensitive fields of > > > >> > reused inodes. > > > >> > (spinlock, next pointer) > > > >> > > > >> Yes, the downside of using SLAB_DESTROY_BY_RCU is that you really > > > >> cannot initialize some fields in the allocation path, because they may > > > >> end up being still used while allocating a new (well, re-used) entry. > > > >> > > > >> However, I think that in the long run we pretty much _have_ to do that > > > >> anyway, because the "free each inode separately with RCU" is a real > > > >> overhead (Nick reports 10-20% cost). So it just makes my skin crawl to > > > >> go that way. > > > > > > > > This is a creat/unlink loop on a tmpfs filesystem. Any real filesystem > > > > is going to be *much* heavier in creat/unlink (so that 10-20% cost would > > > > look more like a few %), and any real workload is going to have much > > > > less intensive pattern. > > > > > > So to get some more precise numbers, on a new kernel, and on a nehalem > > > class CPU, creat/unlink busy loop on ramfs (worst possible case for inode > > > RCU), then inode RCU costs 12% more time. > > > > > > If we go to ext4 over ramdisk, it's 4.2% slower. Btrfs is 4.3% slower, XFS > > > is about 4.9% slower. > > > > That is actually significant because in the current XFS performance > > using delayed logging for pure metadata operations is not that far > > off ramdisk results. Indeed, the simple test: > > > > while (i++ < 1000 * 1000) { > > int fd = open("foo", O_CREAT|O_RDWR, 777); > > unlink("foo"); > > close(fd); > > } > > > > Running 8 instances of the above on XFS, each in their own > > directory, on a single sata drive with delayed logging enabled with > > my current working XFS tree (includes SLAB_DESTROY_BY_RCU inode > > cache and XFS inode cache, and numerous other XFS scalability > > enhancements) currently runs at ~250k files/s. It took ~33s for 8 of > > those loops above to complete in parallel, and was 100% CPU bound... > > David, > > This is 30K inodes per second per CPU, versus nearly 800K per second > number that I measured the 12% slowdown with. About 25x slower. Hi Nick, the ramfs (800k/12%) numbers are not the context I was responding to - you're comparing apples to oranges. I was responding to the "XFS [on a ramdisk] is about 4.9% slower" result. > How you > are trying to FUD this as doing anything but confirming my hypothesis, I > don't know and honestly I don't want to know so don't try to tell me. Hardly FUD. I thought it important to point out that your filesystem-on-ramdisk numbers are not theoretical at all - we can acheive the same level of performance on a single SATA drive for this workload on XFS. Therefore, the 5% difference in performance you've measured on a ramdisk will definitely be visible in the real world and we need to consider it in that context, not as a "theoretical concern". Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx -- 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