RE: [load balancing] The GSLB Page of Shame

From: P T (pt_lbIZZAThotmail.com)
Date: Sun Apr 04 2004 - 14:59:46 EDT


Shawn,

You are right, customers are not naÔve. They are misinformed. In most (or
maybe all) cases the vendors themselves do not know this information. Their
Web sites prove it (even as I type this, the incorrect information is
there). After speaking with people at most every vendor in this space over
the past few weeks (before releasing this paper), I've seen no evidence to
indicate that vendors have intentionally misled customers. They just didn't
know (neither did I, and I _was_ Mr. GSLB at one of the vendors). At some
point I was told the Earth was not flat. I never questioned it, but I never
personally verified it either. At one point I was also told DNS servers pass
lists of A records unchanged, and didn't RR if the TTL was zero. I never
questioned that either, but in this case I inadvertently discovered it was
not true (only to find out that the DNS experts already knew that. Oh well).

The question is not whether customers need "perfect" vs. not quite perfect
GSLB. The question is whether $30,000 GSLB boxes add any value at all in
practical implementations. They had better, as pair of them cost as much as
a new BWM 745Li... but in most cases they in fact don't add any value, and
definitely don't perform as advertized.

So as stated in the paper:

http://www.tenereillo.com/GSLBPageOfShame.htm

customers already have DNS servers. Without buying any GSLB product,
customers have round robin load-balancing at multiple sites that serve
browser based clients, and (with multiple A records used as intended in the
DNS protocol) have basic, fundamental, failover. That is default DNS
behavior/functionality. If one site goes down completely (connection out,
power out, catastrophe, whatever), clients seamlessly go to the another
site, no connection timeouts, no error dialogs, and load sharing is usually
pretty good (though not "perfect").

Enter the $30,000 GSLB box, with features such as RTT or BGP hop count
proximity or load metrics, response time, DNS persistence, cookie
persistence, etc.

Problem is, if customers want to use these features they paid for in the
$30,000 GSLB box, they must give up the basic, fundamental, failover (by
disabling multiple A records).

So ... customers are not being told that if they want to retain the basic
site failover and pretty good load sharing they already have for free, they
cannot use the "better" load balancing features they paid $30,000 for. How
is that not "dire"?

The paper doesn't say customers need perfect GSLB, or the World is ending.
It says customers are, in many cases, wasting their money, because they are
misinformed when spending it.

I don't think this is bad news for the traffic management industry.
Customers either have the $30K to spend or they don't, and every vendor that
makes a GSLB box also makes other stuff. There are lots of better ways $30K
can be spent to improve a Web site. For example, a later generation
SLB/switch, more servers, multi-site synchronization software, or maybe a
NetScaler Web accelerator box!

Pete.

>From: "Shawn Nunley" <shawnIZZATnunleys.com>
>Reply-To: lb-lIZZATvegan.net
>To: <lb-lIZZATvegan.net>
>Subject: RE: [load balancing] The GSLB Page of Shame
>Date: Fri, 2 Apr 2004 20:31:15 -0800
>
>Iíd like to add that, while I agree with the technical descriptions that
>Pete gives in his paper, and I agree that it is difficult to achieve
>perfect
>GSLB with a DNS-based solution, I donít think the problem is nearly as dire
>as stated. The fact is, customers who pay extra for this feature obviously
>see value in it and demand it in the products. Do not believe customers
>are
>naÔve. Most know the edge cases and possible shortcomings that these GSLB
>products have when compared to more perfect solutions. In other words, the
>people who drive the market are not GSLB purists. They seem to be happy
>enough with the products so far because there is consistent growth in that
>market.
>
>Shawn Nunley, CISSP
>Director, Technology Development
>NetScaler, Inc.
>
>-----Original Message-----
>From: Andy Gravett [mailto:agIZZAThtl.uk.com]
>Sent: Thursday, April 01, 2004 6:01 AM
>To: lb-lIZZATvegan.net
>Subject: RE: [load balancing] The GSLB Page of Shame
>
>Hi Guys,
>
>I read your white paper Pete and thought it to be very well written and
>informative, if a little biased towards DNS, I think that what people tend
>to forget is the internet is still like a huge heterogeneous network with
>many different technologies that both help and hinder its growth everybody
>is pushing their technology as the next big thing.
>
> >From the products I have seen GSLB seems in a lot of cases to be an add
>on
>(buzz word), to increase the products marketability, and in most of cases
>these features are Lab tested and work, but in the real world as
>illustrated
>in the paper I can see your point.
>
>But the appliances that do exist that offer this as a feature are usually
>bought/sold as SLB, or as I prefer to call them Traffic Management
>Appliances, for intra-site traffic.
>
>It's like buying a Duck and asking it to Bark!
>
>If site redundancy is required then DNS multiple host 'A' records does seem
>the logical way to go, but not many people buy an appliance to perform all
>the other SLB functions and GSLB is usually an after thought.
>
>my point is that there is a place for these appliances and DNS redundancy
>in
>site infrastructure design and so I don't think that leaning towards one
>technology or another helps, it's a horses for courses approach that wins
>every time in my opinion, but at the moment if site redundancy is the
>topic,
>I would certainly be moving towards Pete and his theories.... but hey this
>is today who knows what tomorrow will hold...
>
>Regards
>
>Andy Gravett
>
>Technical Director
>
>HTLsecure.net
>
>-----Original Message-----
>From: P T [mailto:pt_lbIZZAThotmail.com]
>Sent: 01 April 2004 13:05
>To: lb-lIZZATvegan.net
>Subject: Re: [load balancing] The GSLB Page of Shame
>
>
>Jay, that's just it... there are _no_ workarounds that fix the issues! I
>thought the paper did a pretty thorough job of explaining why the state of
>the art workarounds are not sufficient. Did you read it?
>
>It's not like any one piece of this information is new, but the sum total
>of
>
>it is definitely not widely known within SLB vendors. For example, several
>of them, including the current market share leader, have information on
>their Web sites explaining how they return DNS A records in a list with the
>"best" IP address first. That doesn't work on the Internet! The DNS gurus
>laugh in our faces. (Hey, I didn't know it either, how embarrassing!). It's
>OK to be wrong. As technologists we just learn from mistakes, fix the stuff
>we build, sell stuff that works, right? So GSLB doesn't work as we all had
>hoped, and we can't fix it. There's no sense pointing fingers or lamenting
>about how browser software should change to observe TTLs or SRV records or
>whatever... even if such fixes were released in software today we would
>have
>
>to live with the installed client base for years. Now we know that, and
>move
>
>on. I hope your company, NetScaler, takes such higher ground.
>
>What would really be shameful is for vendors to, after realizing issues
>such
>
>as these, continue promoting features and products that do not work as
>advertised.
>
>
>
>Pete.
>
>
> >From: William Bivens <wjbivensIZZATdigitalme.com>
> >Reply-To: lb-lIZZATvegan.net
> >To: lb-lIZZATvegan.net
> >Subject: Re: [load balancing] The GSLB Page of Shame
> >Date: Wed, 31 Mar 2004 19:38:08 -0500
> >
> >It's a shame the GSLB manifesto couldn't be more neutral, it makes some
> >very valid point. There are several reasons to use GSLB and several
> >workarounds for the specific problems, it just seems that the author
>chose
> >not to highlight or know them.
> >
> >Sincerely,
> >
> >Jay Bivens
> >
> >
> >On Mar 30, 2004, at 2:01 PM, P wrote:
> >
> >>A paper that describes why DNS based GSLB solutions cannot work reliably
> >>for browser based clients.
> >>
> >>†
> >>
> >>http://www.tenereillo.com/GSLBPageOfShame.htm
> >>
> >>†
> >>
> >>†
> >>
> >>†
> >>
> >>Pete.
> >>
> >>†
> >
> >
> >____________________
> >The Load Balancing Mailing List
> >Unsubscribe: mailto:majordomoIZZATvegan.net?body=unsubscribe%20lb-l
> >Archive: http://vegan.net/lb/archive
> >LBDigest: http://lbdigest.com
> >MRTG with SLB: http://vegan.net/MRTG
> >Hosted by: http://www.tokkisystems.com
> >
>
>_________________________________________________________________
>Free up your inbox with MSN Hotmail Extra Storage. Multiple plans
>available.
>
>http://join.msn.com/?pgmarket=en-us&page=hotmail/es2&ST=1/go/onm00200362ave/
>direct/01/
>
>____________________
>The Load Balancing Mailing List
>Unsubscribe: mailto:majordomoIZZATvegan.net?body=unsubscribe%20lb-l
>Archive: http://vegan.net/lb/archive
>LBDigest: http://lbdigest.com
>MRTG with SLB: http://vegan.net/MRTG
>Hosted by: http://www.tokkisystems.com
>
>
>____________________
>The Load Balancing Mailing List
>Unsubscribe: mailto:majordomoIZZATvegan.net?body=unsubscribe%20lb-l
>Archive: http://vegan.net/lb/archive
>LBDigest: http://lbdigest.com
>MRTG with SLB: http://vegan.net/MRTG
>Hosted by: http://www.tokkisystems.com
>
>
>
>
>
>
>____________________
>The Load Balancing Mailing List
>Unsubscribe: mailto:majordomoIZZATvegan.net?body=unsubscribe%20lb-l
>Archive: http://vegan.net/lb/archive
>LBDigest: http://lbdigest.com
>MRTG with SLB: http://vegan.net/MRTG
>Hosted by: http://www.tokkisystems.com
>

_________________________________________________________________
Limited-time offer: Fast, reliable MSN 9 Dial-up Internet access FREE for 2
months!
http://join.msn.com/?page=dept/dialup&pgmarket=en-us&ST=1/go/onm00200361ave/direct/01/

____________________
The Load Balancing Mailing List
Unsubscribe: mailto:majordomoIZZATvegan.net?body=unsubscribe%20lb-l
Archive: http://vegan.net/lb/archive
LBDigest: http://lbdigest.com
MRTG with SLB: http://vegan.net/MRTG
Hosted by: http://www.tokkisystems.com



This archive was generated by hypermail 2.1.4 : Wed Jun 16 2004 - 17:28:58 EDT