Re: OT: Is WAAS that bad?

From: Zack (Doc) <zack_at_tnan.net>
Date: Mon, 3 Sep 2012 12:32:03 -0400

Alex,

Sounds like your SE doesn't know that much about acceleration products (but
that's not uncommon actually, that's why they have experts like Zack Seils
on call)... Fail-to-wire in these type products is a hardware function, and
can operate without even a box present (there was a lovely video on Youtube
of a Riverbed card being used outside the box with just an alligator clip
supplying power for a fail-to-wire demonstration). Problem isolation is a
hard problem no matter which way you are doing it, but at least with inline
you wind up with an extra sniffer (the Riverbeds at least can sniff the
traffic passing through them for troubleshooting purposes).

WCCP is as much a philosophical decision as it is a technical one. But on
the technical front it is HIGHLY dependent on what you have on either side
of the connection. All WCCP does is build GRE tunnels which then become a
virtual in-path for the traffic.

Symmetrical routing is just as important for proper acceleration as it is
for passing through a firewall, for basically the same reason. But, like
firewalls, if you have some form of state-sharing going on between devices,
some asymmetry can be tolerated. The Riverbeds accomplish this with
"Connection Forwarding"... when I last used a WAAS product, they did not
have anything. Again, I've not tried out their 5.0 product; we just did
our RFP last year, so it will be a couple years before we generally
consider another switch in technology.

On Sun, Sep 2, 2012 at 11:49 PM, Alexander Lim <cisco.alexand_at_gmail.com>wrote:

> Hi Travis,
>
> Thanks for the response.
> But I thought regardless using WCCP or in-line, we still need to make sure
> the traffic flows symmetrically in the network.
> My Cisco SE recommends using WCCP instead of inline. He said fail to wire
> will not work if software hangs/crashed. Other than that, it will be more
> difficult to isolate WOC when doing problem determination because all
> traffic pass through the device.
>
> BTW, do you know if there is any blog/web thats compare the app supported
> by Riverbed vs WAAS? I tried googling but couldn't find, may be there isn't
> any.
>
> Regards,
> Alexander Lim
>
> On 3 Sep, 2012, at 7:23 AM, Travis Niedens <niedentj_at_hotmail.com> wrote:
>
> > Having worked with both WAAS and Riverbed, WAAS 5 brings Cisco back to a
> > decent position - they have traditionally been behind on most things that
> > Riverbed has done. Inline versus OOB - depends on the design / traffic. I
> > personally am not a fan of having to use WCCP. With WCCP you need to make
> > sure you have the proper rules on both ends to ensure symmetric
> > optimization. With inline you need to worry about throughput and proper
> > wiring for fail-open.
> >
> > Do keep in mind that there are certain protocols / traffic types that wan
> > acceleration can and cannot optimize. An example of this would be SQL
> > (Oracle, MSSQL, MySql). Both vendors claim a much lower optimization than
> > say CIFS, FTP, etc. Dynamic content is difficult to optimize and you
> really
> > just see the benefits with TCP optimization - even using the compression
> > doesn't do all that much on text outputs from a DB call.
> >
> > T
> >
> > -----Original Message-----
> > From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On Behalf Of
> > Alexander Lim
> > Sent: Sunday, September 02, 2012 4:05 PM
> > To: alexeim73_at_gmail.com
> > Cc: Cisco certification
> > Subject: Re: OT: Is WAAS that bad?
> >
> > Hi Alexei,
> >
> > Based on your exp, does SteelHead have problem with WCCP? And is its
> > transparent mode working as good as tunnel mode?
> >
> > Regards,
> > Alexander Lim
> >
> > On 1 Sep, 2012, at 7:43 PM, Alexei Monastyrnyi <alexeim73_at_gmail.com>
> wrote:
> >
> >> Comparing to Riverbed WAAS is absolutely behind... first hand
> >> experience
> > from rolling out and supporting both on large scale.
> >>
> >> A>
> >>
> >>
> >> On 9/1/2012 3:29 PM, Alexander Lim wrote:
> >>> Hi Experts,
> >>>
> >>> While looking for some comparison between Cisco WAAS and Riverbed
> > SteelHead, I found this article
> >
> http://m.networkworld.com/community/blog/cisco-finally-shows-waas-app-nav-ci
> > s
> > co-live-0. Is it telling the truth? Is Cisco WAAS really that bad? What
> do
> > you guys use in your network?
> >>>
> >>> Thanks for sharing.
> >>>
> >>> Regards,
> >>> Alex
> >>>
> >>>
> >>> Blogs and organic groups at http://www.ccie.net
> >>>
> >>> _____________________________________________________________________
> >>> __ Subscription information may be found at:
> >>> http://www.groupstudy.com/list/CCIELab.html
> >
> >
> > Blogs and organic groups at http://www.ccie.net
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
> >
> >
> > Blogs and organic groups at http://www.ccie.net
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html

Blogs and organic groups at http://www.ccie.net
Received on Mon Sep 03 2012 - 12:32:03 ART

This archive was generated by hypermail 2.2.0 : Mon Oct 01 2012 - 06:40:29 ART