Oracle FAQ Your Portal to the Oracle Knowledge Grid

Home -> Community -> Mailing Lists -> Oracle-L -> Re: "egenera" 10gRAC ? -- Responses

Re: "egenera" 10gRAC ? -- Responses

From: Hemant K Chitale <>
Date: Tue, 07 Nov 2006 21:49:40 +0800
Message-Id: <>

Thank you, all, for your feedback.

 From Mark Teehan :
We've been running RAC on eGenera linux blades for about three years now (in singapore and other locatons)- currerntly we have 10.1 RAC and 10.2 RAC. The new opteron blades, with 32GB of RAM allow you to run quite a number of instances.
The drivers can be a bit tricky as eGenera use their own virtual network interfaces and that may cause some stability issues. But we run a lot of business off eGenera's with no problems.

 From Alex Gorbachev :
I can say that our clients are quite happy with it and we have great experience with that platform. If you are interested in any particular detail, drop me a mail.

 From Matt Zito :

We've got extensive experience with 10gR2 + egenera. Short story: RAC works equally well with egenera as normal hardware. Long story: since the egenera boxes look just like normal servers, there's no advantage either way. Egenera talks a good game about their software helping with RAC, but I've never seen a customer leverage their provisioning or anything like that to help make RAC easier to deploy (in fact, we've had a number of companies come to us to do RAC provisioning on egenera). I have heard people complain that the egenera interconnect is unstable under very high IO loads, but have not seen it myself. My .02 on the matter: Egenera is very expensive compared with what you can get by assembling your own hardware. The commoditization of high-speed compute connects using InfiniBand means that the last argument for Egenera - that you can dynamically create IO connections with high throughput - has gone away. Just go buy your hardware vendor of choice and get Topspin IB from Cisco with their IO allocation technology.

Hemant K Chitale

Received on Tue Nov 07 2006 - 07:49:40 CST

Original text of this message