Beer & Pretzels, Baby!

“If you’re going to use NFS for your enterprise apps, any generic NFS server will do…

– Said no one. Ever.

So it’s well known that Red Hat is a huge contributor to the Open Source community, especially Linux. But it should also come as no surprise that Red Hat also depends on others for certain core components. A great example is the Linux NFS Client stack, which handles everything I/O related from the OS (and kernel) perspective.

Here’s a fun fact: A number of NetApp engineers currently maintain the Linux NFS Client stack. As code and patches are created, they are tested (relentlessly!) against the NetApp NFS server. Once the code/patches/features/fixes are successfully tested, they are submitted to the upstream Linux kernel maintainers.

(I promise I’ll get back to the beer & pretzels.)

To develop each version of Red Hat Enterprise Linux, Red Hat engineering consumes the Linux NFS Client stack as part of the Linux kernel. Red Hat adds their own code and patches, that are also submitted upstream, and then redistributes it in Red Hat Enterprise Linux.

More importantly, Red Hat & NetApp also collaborate on NFS-related activities. A lot. So while other Linux distributions could technically claim that they pull from the same upstream Linux kernel, they can’t claim the same level of collaboration with NetApp on NFS.

They go together like Beer & Pretzels – imagine NetApp Beer paired with Red Hat Pretzels, with a big bowl of Red Hat Pretzel Cheese Dip made with real NetApp Beer!

The Linux NFS Client stack works with any standards compliant NFS server, but the Linux NFS client stack as distributed with RHEL was created in conjunction with NetApp NFS. In other words, they were made for each other. Like Beer & Pretzels.

And if you’re ever in Raleigh, NC, check out the Boylan Bridge Brew Pub, they’ve got GREAT beer & pretzels.

Captain KVM

2 thoughts on “Beer & Pretzels, Baby!”

  1. How come Windows Server 2012 has the fastest NFS server currently shipping then? And AFAIK the only fully supported commercial server NFS4.1 implementation.

    1. Hi TDR,

      Thanks for taking the time to comment and ask questions. If you search for “SPEC, NFS, NetApp”, you’ll find links to NetApp’s world record in 2011. To quote the fine folks at http://recoverymonkey.org, NetApp achieved over 1.5 million SPEC SFS2008 NFS operations per second with a 24-node cluster based on FAS6240 boxes running ONTAP 8 in Cluster Mode. You can see the actual results on the spec.org site here at http://www.spec.org/sfs2008/results/res2011q4/sfs2008-20111003-00198.html. Windows is not listed for NFS or CIFS.

      Additionally, Red Hat fully supports NFS v4.1. pNFS is still in tech preview (until RHEL 6.4 comes out at the end of Feb).

      If you’ve got a link or 2 regarding Windows Server 2012 being the fastest at NFS, I would love to see how they achieved it. (honest, I really would like to see it.)

      thanks,

      Captain KVM

Agree? Disagree? Something to add to the conversation?