Re: 12c grid control
Date: Mon, 1 Jul 2013 09:07:18 -0300
Message-ID: <CAJ2dSGRcrt+X07uL-i0oTXRSnbBSHjZJww6yrL97eicufCvdEQ_at_mail.gmail.com>
Nuno,
I'm only going to address this:
> That is one of the areas I am not happy with EM12c: exactly how to make
> it a HA system?
> Not easy, from what I've seen so far. At least not without some extreme
> complexity added on. Of course, good old "tar cvf" works and will
> always work. But I thought in this day and age something more
> sophisticated could be knocked together, without needing a full-on RAC
> install or an extensive DG setup!
When you talk HA you need to consider all the potential single points of failure and address them accordingly. That applied to the OEM architecture means you need to have redundant connectivity (network, storage, power, etc), redundant database systems for the repository and redundant management servers.
As this is an oracle product, you are stuck using Oracle Database and are thus limited by its HA capabilities, and yes that means RAC and DataGuard. (note that I say AND, because I consider OEM to be a primary system and need the highest possible level of protection). Now, this holds true for any version of OEM, it has always used an Oracle database as repository and the only way to guarantee the repositorie's availability is through Dataguard (redundant hardware, storage and software). As for the OMS, you can install multiple OMSs without much trouble (it took me two or three tries to get it right)... and there you have the entire HA solution with a couple of servers, a decent storage and
I agree that every new version of every tool, especially tools critical to operations must be thoroughly tested, and that the EM12c base release was a bit (lot) buggy and that the BP1 did solve a lot of the bugs but not all of them (far from it) and introduced a few new ones. But this is true for any base release by oracle, at least in the last decade (do I need to remind anyone about 10.1.0.1?).
Now my 0.02 AR$ (which is not worth much, really). Every DBA/Manager/whatever must make their own decision on whether to adopt new technology or not. I always try to stay on the bleeding edge of technology, I like it, I like to learn, I like to test features, I like to think of ways to test a new database or feature and honestly I sometimes just like the challenge. There is, however, a long way from toying around in a lab to implementing a tool or new db version in a productive environment. And if tested properly, most (if not all) bugs that would affect you would be discovered in the testing phase and appropriate actions may be taken (deciding not to implement that version of the product or requesting Oracle Support for a bug analysis and patch for instance).
So in the end this discussion is moot as everyone needs to evaluate the tool based on their own environment and needs. The most balanced answer I can think of to eh OP's question is this:
EM12cCC is a rather new product and as such it's subject to undiscovered bugs. You should thoroughly test its capabilities and ensure all the functionality you are interested in actually works as advertised prior to deciding to implement it. And certainly, you should start with non-productive environments. However there are shops out there that have successfully implemented it and it meets their needs. So there is at least a chance that it will work for you.
sorry for the wall of text.
hth
Alan.-
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Jul 01 2013 - 14:07:18 CEST