|
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Ryan Leathers wrote:
> If you are only doing lvs-nat and you happen to run Red Hat Linux, then
> Piranha is a fine choice. The interface is reasonably intuitive so if
> you don't care to gain a thorough understanding of how lvs works, you
> can likely get away with using piranha to configure it for you. On the
> other hand, if you need to do lvs-dr or use fw marks or anything exotic,
> then piranha is not for you.
>
> On an entirely different note - ultra monkey is a very nice packaging of
> all the things, EXCEPT a gui management interface, that one would need
> for load balancing. I use it on 74 servers (37 pairs) to provide
> fail-over for an application which has no inherent fault tolerance. I
> used to use it for a web farm as well, but I switched to a Red Hat
> conforming LVS along with piranha in order to make it easier for web
> admin types to adjust web farm host details.
>
> On Thu, 2005-11-10 at 11:35 -0700, Dan Trainor wrote:
>
> Graham David Purcocks M.A.(Oxon.) wrote:
>
>>Ultramonkey is a pre-packaged complete implementation of LVS, heartbeat
>>and ldirectord. Which gives you (in order) load balancing, director
>>failover and real server monitoring.
>
>>Piranha is RedHats offerering of the same.
>
>>Keepalived is, I think, equivalent to ldirectord but I'm not sure as I
>>don't use it.
>
>>Hope this helps.
>
>>On Thu, 2005-11-10 at 18:21, Dan Trainor wrote:
>
>>Hello, all -
>
>>Once again I call upon the help of you fine people in helping me better
>>understand exactly what I'm looking at here. Before we get started, I'd
>>first like to thank you all who have helped me in the past. You're an
>>incredible help.
>
>>I've been reading an excellent article by Mr. Zhang on linux-mag.com,
>>http://www.linux-mag.com/2003-11/clusters_01.html. If you have not yet
>>read it, I highly suggest that you do. It is very informative.
>
>>While reading this article, UltraMonkey, Piranha, and Keepalived were
>>briefly mentioned. Although there was a little intro given about all
>>three, their purpose seemed a bit fuzzy to me.
>
>>It seems to me that all three of these services provide the same type of
>>service - they all determine which node is up/working/doing stuff, and
>>deals with this circumstance as it sees appropriate. What I don't quite
>>understand is the subtle differences between the three, or if I'm just
>>completely wrong here. All three describe themselves as dealing with
>>high availability and load balancing, but I can't really find a
>>comparrison between the three.
>
>>If anyone might be able to point me in the right direction, or just give
>>me some links as to where I can read about the differences between the
>>three, I would greatly appreciate it.
>
>>Thanks
>>-dant
>
> Thanks for the reply, Graham -
>
> It doesn't look like there's been much development on Piranha in quite
> some time here. Would it be safe to conclude that UltraMonkey may be my
> best bet here, with consideration given to the fact that I'm quite new
> at this as of yet?
>
> Thanks
> -dant
Thanks for the reply, Ryan -
I do wish to fully understand LVS, and I wasn't asking which was
"easier" to use. As far as I'm concerned, if it's not difficult, I
haven't learned anything.
I'll continue to explore my options here. On top of that, I am leaning
more towards an LVS-DR approach over an LVS-NAT approach, due to the
services which we will be offering.
UltraMonkey seems to be very nice indeed, and that's what I think I'm
going to roll with here. However, I will take everything into
consideration to help me make the best judgement.
Thanks!
- -dant
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
iD8DBQFDc5ythTPx3xy3bu0RAmCyAJ9oAfyjjHxxlN4hevMFhnuhK3vbLwCgpsuJ
Xbbai5i8QguW+g0GA4LLIOc=
=gDKN
-----END PGP SIGNATURE-----
|
