"doubletalk"
Good evening, BD -
Hmmm..."total storage market", eh? You wouldn't mind taking a few minutes to define for us exactly what you mean by that extraordinarily broad term, would you? Thanks in advance!
You mention WAFL.
Regarding WAFL...from NetApp's Web site:
This paper describes WAFLTM (Write Anywhere File Layout), which is a file system designed specifically to work in an NFS appliance. The primary focus is on the algorithms and data structures that WAFL uses to implement Snapshots , which are read-only clones of the active file system. WAFL uses a copy-on-write technique to minimize the disk space that Snapshots consume. This paper also describes how WAFL uses Snapshots to eliminate the need for file system consistency checking after an unclean shutdown.
What NetApp calls "Snapshots" is a great feature, and that functionality is offered by many other vendors these days.
Regarding the importance of performance, that is a critically important aspect, and is a feature that sets NetApp apart from, say, SnapServer.
From NetApp's Web site:
"The Direct Access File System (DAFS) protocol is a high-performance file access protocol designed specifically to use InfiniBand architecture as a standard memory-to-memory interconnect between application servers and shared file storage. DAFS promises to improve the performance and reliability of Internet and enterprise applications in data center environments. The DAFS Collaborative is a group of more than eighty companies that organized last year for the sole purpose of creating this new file-sharing protocol."
Certainly, that is important, but as it notes, more than 80 companies have worked together to develop that standard, so it will not be a specific advantage to NetApp. In any event, it is a feature which can be applied - perhaps even more usefully, to solid-state memory than conventional hard-disk based storage.
Here's where NetApp talks about the performance of WAFL:
3.6. Write Allocation
Write performance is especially important for network file servers. Ousterhout observed that as read caches get larger at both the client and server, writes begin to dominate the I/O subsystem [Ousterhout89]. This effect is especially pronounced with NFS which allows very little client-side write caching. The result is that the disks on an NFS server may have 5 times as many write operations as reads.
( netapp.com
I submit to you that, if you're using a solid-state disk, and thus don't have to concern yourself with the rotation and seek delay which are huge problems with conventional disks, that a system tuned specifically to address those issues isn't as relevant.
Your comment "How important was it in the first place?" (regarding the importance of NetApp's performance advantages) is quite an amazing statement, because high performance is what NetApp filers are about !
Without high performance, the other nice features, such as Snapshot, don't mean mean nearly as much.
Anyway, I find the discussion interesting, if nothing else, and look forward to thoughtful and insightful comments by you and others. As I have no stake in NTAP, EMC, VRTS, SUNW, or any of the other players, (and won't bother for another two years, at least) I'm using this time for learning, and establishing my future strategies.
Wishing everyone a great weekend,
Steve |
