Re: worgroup server vs enterprise server

Doug <ddmc_at_netins.net> writes:

>I would like to get an opinion on the following.
 

>We are looking at putting a database on two unix systems and are
>investigating what database would be best for us. Oracle 7.3.x is one
>of the possibilites.
 

>We need the two databases to stay in sync. When an update happens to
>one, I want the other to reflect the update, so I am looking at
>replication capabilities. I have been talking to a sales rep, but he
>admits that he needs to talk to someone with more technical knowledge
>before he can answer some of my questions.
 

>My understanding is that the workgroup server provides some replication
>capabilities. The price is right for us. The sales rep seems to think
>this version will not be sufficient for our needs, that this will not
>keep the two databases in sync. He feels we will need the enterprise
>server version. This is definitely a pricey alternative to the
>workgroup. He mentions that there would be one common disk the two
>systems would share to pull the info from. I understand what he is
>saying, but that is not what we want. If either one of the unix systems
>crash, all work automatically transfers to the other system. We want
>this system to have all the current data on it so it can continue to
>process the work. I do think we want two databases. Maybe someone can
>explain how a shared database could be utilized if the system it resides
>on crashes.
 

>Hopefully I have given enough info to describe our processing needs.
>Does anyone know if the workgroup server will do what we need? Do we
>really need the enterprise server? I understand that the enterprise
>positions us for future growth. The reality of cost makes it necessary
>to consider the workgroup version if this will truly meet the needs I've
>outlined.
 

>Thanks in advance for any advice you can provide,
 

>Doug

A possibility might be using a Hot Standby database - a separate database on a different machine is kept in sync. with production server up to the last archived redo log, which logs are accessed on the standby machine and applied against the standby database as they are produced. You won't get failover, but you'll get a pretty much up to date database you can switch production over to with a relatively brief period of time. Poor man's OPS if you will. Received on Mon Nov 10 1997 - 00:00:00 CST