ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

Standby server strategy

 
SOLVED
Go to solution
scott_417
Frequent Advisor

Standby server strategy

For a web or a database server either DL360 or DL380, in the case of the crashing, I'll need a standby server. I have some questions about this:

1) What is the general strategy for choosing a standby server? Simply prepare a duplicate/redundant one (exactly same model)?

2) Do I have any other choice, in stead of using DL380 stands by for DL380, or DL360 stands by for DL360?

Thanks to help.
12 REPLIES
Stephen Grimmer
Trusted Contributor

Re: Standby server strategy

Steven Clementi
Honored Contributor

Re: Standby server strategy

Scott:

1. If the business needs warrant it, you should probably consider a cluster. With a cluster, you can run your application and database on separate boxes, but if one of them fails... the cluster automatically restarts the failed services on the other node.

2. You do not necessarily need to have a matching server. You should have something that can handle the full load of the application or database otherwise you will be running in a reduced/degraded state.


Steven
Steven Clementi
HP Master ASE, Storage and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5)
RHCE
NPP3 (Nutanix Platform Professional)
scott_417
Frequent Advisor

Re: Standby server strategy

What is the difference between this Cluster logic and the Oracle's RAC technology? Found very similar, but want to know more major difference.


Thanks


Scott
Steven Clementi
Honored Contributor

Re: Standby server strategy

I can not speak about Oracle RAC as I am not familiar with it as much as i would like to be.


If your using Windows, a Windows Cluster allows your application and/or database to be highly available and redundant. You can lose 1 node of the cluster and still be running your app/db with no issues.

A Windows Cluster would allow you to configure multiple applications and a database.

I beliebe Oracle RAC is only capable of clustering just the oracle piece(s) while I can cluster, in Windows, DHCP, IIS, WINS, File and Printer Sharing, Exchange, SQL, other genric apps.... all on the same cluster.


Steven
Steven Clementi
HP Master ASE, Storage and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5)
RHCE
NPP3 (Nutanix Platform Professional)
Ivan Ferreira
Honored Contributor
Solution

Re: Standby server strategy

The hardware depends of the operating system and the cluster type.

For example, a Microsoft Cluster requires the exact hardware configuration in both nodes.

Linux Clustes does not requries the same hardware configuration.

Oracle RAC is a parallel processing cluster that provides fault tolerance for the database only. An Operating System cluster may provide you failover capabilities for other aplications also.

Oracle, provides "standby database also", that is different than RAC. The "standby database" is a replica of a production database, but if the production database fails, you need to do some job to activate the standby database, the "failover" is not automatic.
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
Steven Clementi
Honored Contributor

Re: Standby server strategy

Microsoft recommends the exact hardware, it does not require it.

Steven
Steven Clementi
HP Master ASE, Storage and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5)
RHCE
NPP3 (Nutanix Platform Professional)
Ivan Ferreira
Honored Contributor

Re: Standby server strategy

I think that the way that they expose the "recommendation" makes you believe that is mandatory.

http://www.microsoft.com/windows2000/techinfo/planning/server/clustersteps.asp

All hardware should be identical, slot for slot, card for card, for all nodes. This will make configuration easier and eliminate potential compatibility problems.
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
Steven Clementi
Honored Contributor

Re: Standby server strategy

Ivan:

I absolutely agree with you. Just saying that it is not a hard requirement.


Steven
Steven Clementi
HP Master ASE, Storage and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5)
RHCE
NPP3 (Nutanix Platform Professional)
Ivan Ferreira
Honored Contributor

Re: Standby server strategy

Steven, I also agree with you. It's just a recommendation, and is up to the desition maker if they use the same hardware or not. I wasn't refuting your post ;).
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
Steven Clementi
Honored Contributor

Re: Standby server strategy

Oh, I know. We just like going back and forth like this cause it is fun. (at least for me it is).


;o)

Steven
Steven Clementi
HP Master ASE, Storage and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5)
RHCE
NPP3 (Nutanix Platform Professional)
Prashant (I am Back)
Honored Contributor

Re: Standby server strategy

There are different options
Like cluster with EVA with mutiple disk arrays.
Or
Concept os sucriber or publisher

Differnet options are there.
MS recomend to have identical piece of h/w on servers
Nothing is impossible
Steven Clementi
Honored Contributor

Re: Standby server strategy

Getting back to the point at hand...


It would probably be nice to get a packaged DL380G4 Cluster with MSA500G2, if you wanted to look into "cluster options".

the packaged cluster comes with everything you need to get started... minus the OS.

2 Servers, 1 Storage Array, 2 SCSI cables and a crossover cable.

Would work nicely for a Microsoft Cluster or even a Linux Cluster.


Steven
Steven Clementi
HP Master ASE, Storage and Clustering
MCSE (NT 4.0, W2K, W2K3)
VCP (ESX2, Vi3, vSphere4, vSphere5)
RHCE
NPP3 (Nutanix Platform Professional)