Tag: Oracle

Real DBAs use Grid Control…

Hopefully the title got your attention. Of course it could have read, “Real Linux Sysadmins use Cobbler and Puppet…”, or any number of comparable statements and products. The point being, there is a gradual evolution in the way we approach tasks and if we don’t move with them we marginalize ourselves to the point where we are so unproductive we cease to be of use.

A few years ago I was doing a lot of Linux installations and I got sick of running around with CDs, so started doing network installations to save time. I’ve been doing loads of installs on VMs at home recently, so I started doing PXE Network Installations, which saved me even more time. As a result of the article I wrote about that, Frits Hoogland pointed me in the direction of Cobbler, which makes PXE installations real easy (once you get to grips with it). I’m not a sysadmin, so why do I care? Even when I’m installing and running a handful of VMs at home I can see productivity gains by using some of these tools. Imagine the impact in a data-center!

So back to Grid Control. Does anyone remember the days when you kept a “tail -f” on your alert log? At one site I used to have a CDE workspace on an X station just running tails. Then the number of instances got too big, so I used to scan through the alert logs each day to look for issues. The next step was to use shell scripts to check for errors and mail me. This was a pain at one site where I was using Solaris, HP-UX and Windows, which meant I needed three solutions. Then the Oracle 9i Enterprise Manager with the Management Server came into my life. All of a sudden it could manage my alert logs and I could assume everything was fine ( 🙂 ) unless I got a notification email. This feature alone sold me on the 9i management server.

Back then, being a DBA and admitting using Enterprise Manager was a little like announcing to the world you were into cross dressing. 🙂 Time has moved on, the product name has changed and so has its functionality, but essentially it’s still doing the same thing, which is reducing the effort needed to manage databases (and other things). The difference is that rather than managing 40 instances, teams are now managing thousands of instances.

Of course, none of this is new. I guess it’s just been brought into focus by a few things that have happened to me recently, like the PXE/Cobbler thing, the recent demise of my Grid Control VM at home and the constant talk of cloud computing and SaaS etc.

Specialists and performance consultants have the time to obsess over minute detail. The day-to-day DBAs and sysadmins have to churn through work at a pace, with reliable and reproducible results. Failing to embrace tools, whatever they are, to aid this is career suicide.

Cheers

Tim…

Grid Control (10.2.0.5) on OEL 4.8 x86_64…

I recently managed to screw up my Grid Control VM, so I was forced to reinstall it. I’d not written an article on it since 10gR2, as I’ve just been applying patches since then. Faced with a new install I decided to go x86_64 and do clean install direct to 10gR5.

I used OEL 4.8 as it is the highest supported OS version. I went with the standard new database installation as I can’t really be bothered messing about with an 11g database as the repository. I think of Grid Control in the same way I think of Oracle Apps. They are shrink-wrapped products and I try to keep them as basic as possible. Anything that can present a problem, usually will.

I must admit to liking Grid Control as a product, but the installation and configuration is truely horrible. At every step of the way you are waiting for the next disaster. There are just too many working parts and it eats resources like nobodies business. Am I the only person who thinks the infrastructure is overly complicated for the job it is actually doing? I really hope the 11g version (if it is ever released) will be neater.

Cheers

Tim…

Oracle Database File System (DBFS) in 11gR2…

I’ve been having a play with the DBFS functionality in 11gR2.

For the most part it is extremely simple. If you are working on Linux then it’s a really neat solution. On any other platform you are limited to using the dbfs_client tool, which is currently lacking a lot of functionality. Even so, it’s a good first step.

I am having a bit of trouble with the “/etc/fstab” mounting. I’ve included it in the article, with a warning that it isn’t working for me. If anyone has got it to work I would really appreciate some input on what I’ve got wrong. The documentation is a little sparse on this subject even with some digging around it’s proving difficult.

I’ll probably be witing another article on a different aspect of DBFS soon.

Cheers

Tim…

Magic Pixie Dust Recovery Area…

This morning, as I was scanning through the 11gR2 version of the Backup and Recovery User Guide and I saw the term Fast Recovery Area. My first thought was, “Isn’t that meant to be Flash Recovery Area?”. This was quickly followed by that flash (excuse the pun) of embarrassment when you think maybe it changed ages ago and you are the only person in the world that didn’t notice… 🙂

Well it seems in 11gR1 it was still called the Flash Recovery Area, but 11gR2 has replaced the word Flash with Fast. I guess the name change makes sense as the word Flash may fool people into thinking it’s just for Flashback stuff.

I’m sure in 12g it will be called the Magic Pixie Dust Recovery Area…

Cheers

Tim…

DNS Configuration for SCAN and Editions…

A couple of new articles have crept out recently. The first is me pretending to understand DNS.

I used this configuration in place of the “/etc/hosts” in my VMware RAC installation and it worked great.

The second is a brief romp through edition-based redefinition.

This article started to get really big and feel like a rewrite of the manual, so I stripped most of it out and really just left a couple of examples of how it can be used. I figure this is enough to give you a feel for what it can do, but isn’t as daunting as working through the manuals if all you want is a quick taste.

I’ve seen edition-based redefinition described as a killer feature, but I’m not so sure myself. Don’t get me wrong, I think it is really cool, but “really cool” doesn’t always become “frequently used”. As I was playing with it I had flashbacks to Workspace Management introduced in 9i. I’ve spoken to a lot of poeple over the years and very few even remember it exists, let alone use it.

There is nothing conceptually difficult about edition-based redefintion, but there are potentially a lot of working parts involved and therefore a lot of scope for human error and/or confusion. I’m sure some people have been praying for something like this for a long time, and others will remain blissfully ignorant of it forever. It would be interesting to gaze into a crystal ball and see how much this stuff is used in a few years time (and get some lottery results).

Cheers

Tim…

Something old and something new…

A couple of articles crept out of the last couple of days.

First a revamp of an old article about XML over HTTP to bring it in line with some of the stuff I’ve been presenting recently.

Next, something new that caught my eye to do with auditing in 11gR2.

I know it’s hard to get excited about auditing, but this stuff is pretty neat.

Cheers

Tim…

Oracle 11gR2 on Fedora 12…

I had a play around yesterday and installed Oracle 11gR2 on Fedora 12:

I think this might be the last time I install Oracle on Fedora. Why? For two reasons:

  1. When I first started doing installs on Red Hat 7 (pre Enterprise Linux) it was the only sensible choice. After the introduction of Enterprise Linux there were no free alternatives, so installing on Red Hat Linux, then Fedora Core and now Fedora seemed a viable alternative to paying for an Enterprise Linux distribution. Fast forward a few years and we have Oracle Enterprise Linux which is a supported binary clone of RHEL and most importantly it’s free, so that seems like the logical choice for testing installations.
  2. When I started doing these installations I had to swap hard drives, so having an installation that ran on my Linux desktop was import to me. Once again, fast forward a few years and I never run Oracle directly on my desktop or laptop OS. I always use a Virtual Machine and install Oracle on Oracle Enterprise Linux. Since there are a number of free virtualization products available, there really is no barrier to entry here also.

So why did I bother with this installation? It seems that Fedora12 will likely be the base which RHEL6 is built on, so it is mildly more interesting to me that previous Fedora releases.

Anyway, the articles are there, but do I care about them? No. I’ve always said I write about what interests me and Fedora (or any other Distro for that matter) are completely irrelevant to me now. I see no point in installing Oracle on anything other than RHEL, and really by that I mean Oracle Enterprise Linux, so that is likely to be what I do from now on. Of course, you should never say never. 🙂

For those people banging their heads against a brick wall trying to install Oracle on unsupported distributions, I say download a free virtualization product (VMware Server or VirtualBox) and use it to install Oracle Enterprise Linux and use that for all your future Oracle installs. It’s free and easy.

Cheers

Tim…