|
On Tue, 26 Sep 2006, Dave Augustus wrote:
Hello Joe,
No, you don't know me and I don't know you. I have been on this list for
many years and you and others have helped me many times. :) My
involvement with Linux goes back to 2000 with Redhat 6 and I am an avid
promoter of open source software.
see http://archive.linuxvirtualserver.org/html/lvs-
users/2002-12/msg00103.html
We currently have a 6 node LVS in production- since 2002 I might add.
By opportunity, I don't mean to sound threatening. We have been using
LVS since 2002. And it has worked great. My goal is to help the project
-to give back to the project.
The opportunity exists in the potential of getting Seibel to validate
LVS as a 3rd party source of load-balancing.
This may be fruitless but the *positive* side of this would be:
- the attention that it gives the LVS project.
- the savings to companies in not having to purchase other products
- recognition that LVS is ready for prime time
It's got my vote. I work for an ISP and we use LVS (almost) exclusivly for
loadbalancing.
6 active/passive director pairs (with hearbeat for failover) balancing for
some ~40 real-servers.
Aside from one problem balancing UDP (for DNS servers) it's never failed
us. That problem was that the connections where not expired, and it would
run OOM and crash unless we did a preemptive failover & reboot. (Which
meant we could still run the service without 'customer impact'.)
That problem was fixed in approx 2 weeks or so after my initial report
here on the list. Fastes BUG-fix I've ever had, from any 'vendor'!
+It even got me my name in the kernel changelog (2.6.12-rc5 IIRC) :)
We started using LVS because we couldn't settle the argument over who
should maintain de loadbalancers, us - the sysamins of the stuff behind
them - or the network dept who thought they should do it because it was
networking equipment....
So we (the sysadmins) made a "business case" for LVS by exagerating some
requirements we knew they could not meat with the (then) current
loadbalancers and could be met, easy and cheap, by using LVS. Management
bought into it and we started (carefully) introducing LVS. We've never had
any reason to consider using anything else for loadbalancing ever since.
Thanks to all who contributed.
Regards,
Mark.
I realize that this may or may not be what you as a project leader want.
Help me out here.
Thanks,
Dave Augustus
On Tue, 2006-09-26 at 12:31 -0700, Joseph Mack NA3T wrote:
On Tue, 26 Sep 2006, Dave Augustus wrote:
Hello All,
Due to the awesome product that LVS is I have obtained the
support of management to deploy LVS as the load-balancer
for our new CRM deployment.
I'm glad you like it and I hope it's being useful.
Although LVS was started initially by the bright ideas and
hard work of one person, it's now a distributed development
project where people do as they see fit. There isn't a
spokesperson for LVS and I'm speaking personally here. I'm
rather wary when someone who I don't know comes bounding
over to tell me that something I don't know about will be a
great opportunity for me. As well you aren't clear about
whether you want help from us or are just keeping us
informed of what you're doing.
We are considering approaching Siebel to get their
approval and support of using LVS instead.
I don't know anything about CRMs although I know people who
use them. Does Siebel understand the GPL? Do you?
This appears to be a great opportunity for this project.
CRMs weren't on our list of things to do. I suspect that
there's nothing unique about CRMs - load balancers have no
idea what sort of application they're loadbalancing. Perhaps
you could fill us in here.
Joe
_______________________________________________
LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
or go to http://www.in-addr.de/mailman/listinfo/lvs-users
|
