On Wed, 2005-02-09 at 20:15 +, Gary Stimson wrote:
> On Wednesday 02 February 2005 23:38, Jens Vagelpohl wrote:
> > On Feb 3, 2005, at 0:03, Miguel Ruiz Velasco Sobrino wrote:
> > > --- [EMAIL PROTECTED] wrote:
> > >> The DNS approach does not handle single or multiple system failures,
> > >>
On Wednesday 02 February 2005 23:38, Jens Vagelpohl wrote:
> On Feb 3, 2005, at 0:03, Miguel Ruiz Velasco Sobrino wrote:
> > --- [EMAIL PROTECTED] wrote:
> >> The DNS approach does not handle single or multiple system failures,
> >> only very elementary load balancing over a lengthy period of time.
> Backhoe's are pretty indiscriminatethey'll cut copper just as easily as
> fiber.
Not really. They tend to do a lot more damage to copper, because there
are usually a ton of conductors in the copper cable, and the bundle may be
strong enough to be pulled out of the ground. This causes all
Jon Bebeau wrote:
Mark,
I've been following this thread with some interest as we're gearing up
for load/failover processing. Can you elaborate on the garp and IP
takeover process, like what software packages do that in Linux or point
me to a site for more info?
Thanks,
Jon
Jon,
http://www.linu
m: Joe Greco <[EMAIL PROTECTED]>
Subject: Re: [Asterisk-Users] Re: load balancing 20 asterisk servers
To: asterisk-users@lists.digium.com
Message-ID: <[EMAIL PROTECTED]>
Content-Type: text/plain; charset=us-ascii
> I'm trying to stay away from a software based load balancer cause w
On Feb 3, 2005, at 17:08, Jon Bebeau wrote:
Mark,
I've been following this thread with some interest as we're gearing up
for load/failover processing. Can you elaborate on the garp and IP
takeover process, like what software packages do that in Linux or
point me to a site for more info?
http://
From: "Mark Musone" <[EMAIL PROTECTED]>
To: "Asterisk Users Mailing List - Non-Commercial Discussion"
Sent: Thursday, February 03, 2005 10:36 AM
Subject: Re: [Asterisk-Users] Re: load balancing 20 asterisk servers
Don't confuse load balacing with failover
> Don't confuse load balacing with failover. They are quite different
> beasts and are handled differently. (sure, they can be "combined" into
> one solution, but they are still effectively very different)
>
> Round Robin DNS based load balancing is still a viable load balancing
> solution (read s
Nortel AAS-2000 range of LB's can do this today.
Giles
- Original Message -
From: "Patrick" <[EMAIL PROTECTED]>
To: "Asterisk Users Mailing List - Non-Commercial Discussion"
Sent: Thursday, February 03, 2005 12:38 PM
Subject: Re: [Asterisk-Users] Re: loa
Don't confuse load balacing with failover. They are quite different
beasts and are handled differently. (sure, they can be "combined" into
one solution, but they are still effectively very different)
Round Robin DNS based load balancing is still a viable load balancing
solution (read some of the e
2005 9:32 AM
Subject: RE: [Asterisk-Users] Re: load balancing 20 asterisk servers
-Original Message-
From: Rich Adamson [mailto:[EMAIL PROTECTED]
You have fiber-seeking-backhoes in your area? Wow!
They're everywhere, man! When I was in college an entire nearby town lost
all ph
> -Original Message-
> From: Rich Adamson [mailto:[EMAIL PROTECTED]
> You have fiber-seeking-backhoes in your area? Wow!
They're everywhere, man! When I was in college an entire nearby town lost
all phone service for 24 hours due to a backhoe cutting a fiber optic cable.
3,000 people wit
Rich Adamson wrote:
[snip]
I'm not aware of any balancers that
can do that today.
Afaik Cisco is working on SIP aware loadbalancer functionality. Don't
know what the status is and since it's Cisco I'm sure it will cost a bundle.
Regards,
Patrick
___
Ast
I'm trying to stay away from a software based load balancer cause what
happens if that server fails?
Its far less likely for a piece of dedicated hardware to fail than an
actual
computer.
"A piece of dedicated hardware" runs an OS as well.
I've been running "software" solutions for virtually every
> DNS based load ballancing has it's place, as dose using an
> application level switch.
>
> Say an earthquake takes out your California data center.
> Shortly the DNS servers will notice and pull that center's
> record. However do to caches and all this is not fast
> and users will notice.
>
> How about a management server that polls the asterisk servers every
> minute with snmp to check cpu and ram cache, maybe even drive space.
> Then you could have a script decide whether the server can handle
> anymore connections.
There are lots of different ways to measure how busy a server
> >> The DNS approach does not handle single or multiple system failures,
> >> only very elementary load balancing over a lengthy period of time.
> >
> > Are you shure of that? I'm aware that the load criteria is trickier,
> > but very possible.
Yes, very sure. Look at past posts relative to the
On Feb 3, 2005, at 4:20, Matthew Boehm wrote:
I'm trying to stay away from a software based load balancer cause what
happens if that server fails?
Its far less likely for a piece of dedicated hardware to fail than an
actual
computer.
There are useful things like "heartbeat" which can transparently
> I'm trying to stay away from a software based load balancer cause what
> happens if that server fails?
> Its far less likely for a piece of dedicated hardware to fail than an actual
> computer.
You really ought to open up one of those pieces of dedicated hardware
sometime and see what's inside.
> LVS is a single point of failure, but probably so is your router/switch.
> Consider the case where the LVS *is* the router, use good quality
> components for the PC (we should all know about this part on this list),
We've used the via-based eden motherboards for this sort of thing - rock
solid,
On Wed, 2005-02-02 at 21:20 -0600, Matthew Boehm wrote:
> I'm trying to stay away from a software based load balancer cause what
> happens if that server fails?
> Its far less likely for a piece of dedicated hardware to fail than an actual
> computer.
There are many ways to accomplish this, and th
uot;Asterisk Users Mailing List - Non-Commercial Discussion"
Sent: Wednesday, February 02, 2005 5:54 PM
Subject: Re: [Asterisk-Users] Re: load balancing 20 asterisk servers
>
> How about a management server that polls the asterisk servers every
> minute with snmp to check cpu and ram c
DNS based load ballancing has it's place, as dose using an
application level switch.
Say an earthquake takes out your California data center.
Shortly the DNS servers will notice and pull that center's
record. However do to caches and all this is not fast
and users will notice.
What the switc
How about a management server that polls the asterisk servers every
minute with snmp to check cpu and ram cache, maybe even drive space.
Then you could have a script decide whether the server can handle
anymore connections.
I am still a beginner so I am not sure how you could have asterisk
de
On Feb 3, 2005, at 0:03, Miguel Ruiz Velasco Sobrino wrote:
--- [EMAIL PROTECTED] wrote:
The DNS approach does not handle single or multiple system failures,
only very elementary load balancing over a lengthy period of time.
Are you shure of that? I'm aware that the load criteria is trickier,
but
--- [EMAIL PROTECTED] wrote:
> The DNS approach does not handle single or multiple system failures,
> only very elementary load balancing over a lengthy period of time.
Are you shure of that? I'm aware that the load criteria is trickier, but very
possible.
If you use DDNS (dynamic DNS) using B
The DNS approach does not handle single or multiple system failures,
only very elementary load balancing over a lengthy period of time.
> You may want to consider a simpler aproach, why don't you balance the load
> via DNS?
> If you put in a zone file various A records f
You may want to consider a simpler aproach, why don't you balance the load via
DNS?
If you put in a zone file various A records for the same machine, but with
different
IP's, BIND will catch the trick and send a different IP (from the pool yo
defined) each
time a DNS request arrives. That's a si
28 matches
Mail list logo
