Re: Varnish minus VCL?

2007-09-26 Thread Luke Macpherson
> Not to take away from what sounds like a fun and sensible idea, but done 
> correctly,
> the suggestions that Ask made (perbal, or an LB) don't have to be viewed as 
> "bottlenecks"
> or a single-point of failure.  Load balancers can be very good things to 
> have. :)

I kind-of agree. Obviously I want load balancing, it's just I don't
want to have to insert yet another layer to get it.

If we add another layer, that layer has to be provisioned to cope with
the full bandwidth of the backends, plus we need enough spare machines
at that layer to provide failover. If we instead achieve the same
thing using the front-end servers running varnish, we save ourselves
all of the hardware that would be required for the load balancers.

Meanwhile adding load balancing to varnish should have little or no
measurable performance impact, and does not require any additional
hardware either. Moreover, load balancing performance automatically
scales as the number of varnish caches is increased.

So anyway, that's a slightly more in-depth explanation for why I I
think it is good practice to avoid introducing an additional
load-balancing layer.
___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev


Re: Varnish minus VCL?

2007-09-26 Thread john allspaw
Not to take away from what sounds like a fun and sensible idea, but done 
correctly, 
the suggestions that Ask made (perbal, or an LB) don't have to be viewed as 
"bottlenecks"
or a single-point of failure.  Load balancers can be very good things to have. 
:)

-j

- Original Message 
From: Luke Macpherson <[EMAIL PROTECTED]>
To: Ask Bjørn Hansen <[EMAIL PROTECTED]>
Cc: varnish-dev@projects.linpro.no
Sent: Wednesday, September 26, 2007 5:29:12 PM
Subject: Re: Varnish minus VCL?

> How many servers do you have that this is a concern?

Enough. Though frankly, even if you had a small number of servers it
would still be nice to support pluggable load balancing and failover
models.

> Have you considered just putting perlbal or a load-balancer appliance
> between varnish and your backend servers?  Then the dedicated device
> can take care of load-balancing, fail-over and all that fun - and you
> can use the management systems there to take backend servers in and
> out.   Varnish then will just have to do what it does so well,
 caching.

Inserting a bottleneck and single point of failure doesn't sound like
a good idea. Having m varnish servers talking to n backends gives a
much better failover and redundancy model.

By the way, I'm not asking anyone to implement this for me. I am a
software engineer (not a network administrator), and since I have to
do this anyway it's worth doing it in a way that can potentially
contribute back to the project.
___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev





   

Yahoo! oneSearch: Finally, mobile search 
that gives answers, not web links. 
http://mobile.yahoo.com/mobileweb/onesearch?refer=1ONXIC
___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev


Re: Varnish minus VCL?

2007-09-26 Thread Luke Macpherson
> How many servers do you have that this is a concern?

Enough. Though frankly, even if you had a small number of servers it
would still be nice to support pluggable load balancing and failover
models.

> Have you considered just putting perlbal or a load-balancer appliance
> between varnish and your backend servers?  Then the dedicated device
> can take care of load-balancing, fail-over and all that fun - and you
> can use the management systems there to take backend servers in and
> out.   Varnish then will just have to do what it does so well, caching.

Inserting a bottleneck and single point of failure doesn't sound like
a good idea. Having m varnish servers talking to n backends gives a
much better failover and redundancy model.

By the way, I'm not asking anyone to implement this for me. I am a
software engineer (not a network administrator), and since I have to
do this anyway it's worth doing it in a way that can potentially
contribute back to the project.
___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev


Re: Error

2007-09-26 Thread Poul-Henning Kamp
In message <[EMAIL PROTECTED]>, "sure
ndar p" writes:

>
>Hi,
>When i compile the code,i got this error
>"Expected positive indentation".
>
>what is the actual error it is?

file and line information, please ?

-- 
Poul-Henning Kamp   | UNIX since Zilog Zeus 3.20
[EMAIL PROTECTED] | TCP/IP since RFC 956
FreeBSD committer   | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev


Error

2007-09-26 Thread surendar p
Hi,
When i compile the code,i got this error
"Expected positive indentation".

what is the actual error it is?

Regards
Surendar
___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev


Re: Varnish minus VCL?

2007-09-26 Thread Ask Bjørn Hansen

On Sep 25, 2007, at 18:46, Luke Macpherson wrote:

> I'm still kicking around ideas here. I'm trying to think of a solution
> which is workable in the presence of larger numbers of backends, [...]

How many servers do you have that this is a concern?

Have you considered just putting perlbal or a load-balancer appliance  
between varnish and your backend servers?  Then the dedicated device  
can take care of load-balancing, fail-over and all that fun - and you  
can use the management systems there to take backend servers in and  
out.   Varnish then will just have to do what it does so well, caching.



  - ask

-- 
http://develooper.com/ - http://askask.com/


___
varnish-dev mailing list
varnish-dev@projects.linpro.no
http://projects.linpro.no/mailman/listinfo/varnish-dev