Re: [gentoo-cluster] gentoo 2006.1 + openmosix-sources-2.6.12.r577 + openmosix-user?
hello Ramon (cannot be agreed that grids needs to be explicitly defined, but we can leave it as is – that's not important actually) Your example is quite competitive… I think that it works well when properly used. btw I'm working at dell emea :) - glad to hear of dellserv using ;) Let me please explain my conf. Well, we are using 3dns/bigip for web/application/etc layers. I'm expert here so there are no questions. The question is all about of oracle rac that uses LVS in round robin (LVS is configured directly on database nodes). The actual question Our development and app management is divided on two parts: the database and the rest. Oracle/linadm are using LVS when they are building databases for us. (just to clarify the question) Oracle RAC consists of several Active/Active nodes, each of them has its own node address. There is also dataguard but it is out of scope in this context. So, we have, say, N nodes and N node IPs. Normally, these IPs I should use in Oracle client TNS name on app boxes. But linux gents are building LVS on top, where we additionally have K LVS' IPs (K=N) and they are suggesting to use LVS' instead of Oracle's ones. That's probably is fine. Oracle clients on each app box is configured in LB mode. Transactions are short lived then connection is recycled (in a test period I'm speaking of – immediately) So then, I'm using MRTG to see database load…. Yes! It is not loadbalanced properly over time. When I'm injecting F5 that points to N database nodes in roundrobin and proposes 2 BigIP's IPs in the same OraClient configuration it is loadbalanced properly. The question is – I'm trying to find at least 1 idea on why we need to use LVS here. -- gentoo-cluster@gentoo.org mailing list
Re: [gentoo-cluster] gentoo 2006.1 + openmosix-sources-2.6.12.r577 + openmosix-user?
momentics wrote: Btw � is anybody can say anything good about LVS? Our *nix/DBAs are using this with oracle, but seem it works poorly. I designed my last project to use F5 networks device + extended monitor. Does anybody use LVS in grids? I'm not sure, since I'm not sure about the precise meaning of a grid, it seems to be a term used for a lot of different things. However I am running a 70 node webfarm and a 50+ node database farm behind a couple of LVS based load-balancers with keepalived as management/monitor/checking software. I will say very nice things about it, most importantly it has proven to be a very flexible load-balancer setup which configured correctly is able to sustain a massive amount of load. The load-balancers are running on commodity dell hardware. Right now it's load-balancing our 18M pageviews/day website without a hitch. WRT databases, we load-balance a number of mysql-replication setups through it and that is working very good for us. However that is also caused by the fact that connection initiation for mysql is extremely fast and low on resources. Oracle is far slower and resource-hungry wrt to setting up a connection. If you have any questions I'll be happy to answer. Ramon -- gentoo-cluster@gentoo.org mailing list
Re: [gentoo-cluster] gentoo 2006.1 + openmosix-sources-2.6.12.r577 + openmosix-user?
I'm all about of Fault Tolerant systems… But definitely you need a queuing system for your toy - Hanni suggested one - I will try also :) generally It should work well even in your developed code (I'm thinking of simplified development since – ie of selfish/byzantine node behavior in FT space) For all the rest hpc-howto is a right choice (please review a cosmetic bug for latest glibs for this document I sent you before), even in load balanced service oriented grids. hum, can we think of virtual server as for LB for the grid in a manner for the hpc platform? // Btw – is anybody can say anything good about LVS? Our *nix/DBAs are using this with oracle, but seem it works poorly. I designed my last project to use F5 networks device + extended monitor. Does anybody use LVS in grids? -- gentoo-cluster@gentoo.org mailing list
Re: [gentoo-cluster] gentoo 2006.1 + openmosix-sources-2.6.12.r577 + openmosix-user?
Hanni Ali wrote: ok so follow the hpc how to: http://www.gentoo.org/doc/en/hpc-howto.xml but modify for diskless which you seem to have got the hang of. FWIW, I've got plans to put together a doc for a clean setup of a diskless cluster. I'm unhappy with our clustering documentation, although it is better than nothing. Once I finish the LTSP port to Gentoo (which starts as standard diskless nodes), I'll probably write this doc up. Thanks, Donnie -- gentoo-cluster@gentoo.org mailing list
Re: [gentoo-cluster] gentoo 2006.1 + openmosix-sources-2.6.12.r577 + openmosix-user?
On Wed, 20 Dec 2006 09:50:29 + Hanni Ali [EMAIL PROTECTED] wrote: I recommend a 'generic cluster' with a good queueing program like DrQueue to achieve similar results to OpenMosix. Hanni ok, thanks. ; ) i am looking into generic clustering. I want to thank all of you for your kind assistance, it's really appreciated. nobody's really working on this on the forums. -- gentoo-cluster@gentoo.org mailing list
Re: [gentoo-cluster] gentoo 2006.1 + openmosix-sources-2.6.12.r577 + openmosix-user?
On Tue, 19 Dec 2006 02:37:25 +0300 momentics [EMAIL PROTECTED] wrote: Łukasz [EMAIL PROTECTED] wrote: openMosix 2.6 is still in development, and no user tools are available yet. You must not compile 2.4 source w/gcc-4, it works with gcc-2/3 only. And if you start from stage, choose no-nptl version as nptl is not supported in 2.4, but is in glibc-2.4 which is not downgradeble. Make sure to downgrade couple things like kernel headers after choosing 2.4 profile. however that's an error for generic clustering that is fixed in HPC HOW-TO CVS: Clear-Text: http://bugs.gentoo.org/show_bug.cgi?id=158491 Secure: https://bugs.gentoo.org/show_bug.cgi?id=158491 thanks for that link!! do y'all suggest 'generic clustering' nowadays? -- gentoo-cluster@gentoo.org mailing list
