Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: HP MC / Service Guard / Oracle 8i
Christian Hartmann wrote:
> On Thu, 27 Feb 2003 11:02:21 -0800, "oracle1"
> <santyspamsharma_at_hotmail.com> wrote:
>
>
>
>>yes, you got that right finally. A package in HP term is container for all >>resources that a service needs. So in case of database, services to monitor >>are instance ( PMON or SMON) and listener while resources are datafiles, >>init, control, redo. Mail services, NFS are some other serices that service >>guard is usually used for.
simple, what if the primary machine goes down due to an hardware failure ? like memory error, cpu issues etc. then you have secondary machine to fail over the databases. basically mc/sg supports to an extent of hardware redundancy. however that being said, if the 'shared disk array'
fails (bad disks..corruptions) then mc/sg doesn't come for rescue.
Now if you had a standby database on a separate box, then your database should have been up-to-date (lagging at least by the current redolog file of the primary database that were not archived).
>
> If I use a Bea application-server (Weblogic) with a connection-pool of
> already established database-connection to the first instance, I
> assume that if the first instance brokes down I have to restart the
> Bea Weblogic, don't I? Or wil the open connections of the WebLogic to
> the oracle-database been hold?
>
yes. you have to restart you apps, since mc/sg shuts the active database
with 'shutdown abort'...halts the package...switch the package to fail
over node...starts package..starts database.
> Regards,
>
> Christian
>
-- -StanReceived on Fri Feb 28 2003 - 02:09:29 CST