Another great trip north of the border to enjoy a great meeting with a great group of Oracle customers
... nothing so enjoyable as traveling by seaplane ... nothing so enjoyable as having some decent scotch with good friends
... nothing so enjoyable as visiting a civilized society.
VicOUG President Caleb Small
Camosun College Meeting Room before the meeting started
Friday 14 September, 2011
It has been a long time since I've been inclined to write anything here.
But today was a great day among friends in Salt Lake City at the Utah Oracle Users Group Fall Seminar.
You can find my papers on the Presentations page and here are three of my favorite pictures from today.
UTOUG President Michelle Kolbe
Phil Rizzo - Oracle Golden Gate
Friday 13 May, 2011
Anything but a stereotypical Friday the 13th.
All is going well Yesterday Melina took Francisco and me to Sandanski: A Four hour trip through beautiful country with a side-trip to Melnick:
Home of the best wine-vinegar on the planet (alas they think it drinkable wine) but still a very nice trip to a city that appears to have
reached its zenith in somewhere around the year 1230.
Today was the first day of the conference and I presented a new paper on PL/SQL Tips and Tricks which was a way of packaging Oracle 9i to 11g
new features in a way that didn't say ... "hey you should have learned this stuff years ago." Well received with lots of questions.
Tonight was the first conference dinner ... great food and the pictures will tell the rest of the story except that Joze and his wife, Kusai Mensah,
and Chris Date joined us for dinner.
I am getting ready to for a mental health break the Netherlands, Germany, Bulgaria,
and Finland with a possible side-trip to the UK. Here is the link for the trip and any resulting images
Tuesday 22 February, 2011
An Incredible Lesson in Humility
Saturday, January 10, 2011
Nothing like a good rant to get the keyboard moving.
So if you are unprepared for me dumping on network administrators please go visit a porn site as it will be a far less offensive.
Over my 30+ years in IT there is one thing I have learned is a constant. Not the value of Pi, not the number of atoms in a mole,
not the speed of light in a vacuum: Rather it is the ability of network admins to state, authoritatively and with a straight face, that the network is fine.
Their concept of fine seems to roughly equate to the fact that the hardware is plugged in, the little green link lights are blinking,
nothing red is lit up, and the fire alarm is not sounding.
If databases were run with the same degree of intelligence and attentiveness as a network router we would log in as SYS, type SELECT * FROM dual,
and if we did not get an exception, declare everything was perfect. Were it not for what just happened in Arizona, so I am choosing my words with care,
I might be inclined to recommend that we plug a couple of them in to their own networks and see if they light up.
Why this rant? Because at one of the largest telecoms on the planet, sorry no names, I have been monitoring the cache fusion interconnect on a four node
cluster since June of 2010. And since at least July of 2010 stating that there is a physical problem with the network hardware.
Was I greeted by someone eager to jump in and identify the obvious problem, well supported with latency data? If you think so I recommend Thorazine.
Of course not. I was assured over and over and over and over again that everything was fine.
Fine that is until it broke. Broke as in I spent 5 days over what was supposed to be a Christmas holiday, with family,
babysitting an outage bridge of which the end results were that two IO cards and one NIC card were replaced along with one switch port: All defective.
And then was treated to a wrap up in which an admin admitted that there was evidence in the logs that some of the hardware had been bad since October.
That is three months during which they claimed it had to be a database issue and I strongly suspect if they looked more they'd find the hardware was
bad a lot longer than that.
So the lesson learned, or perhaps I should say relearned, is to never trust a network admin's word to be worth more than the energy it takes to dismiss
it unless that network admin can back up their statement with physical evidence ... not their usual absence of evidence. </RANT>
Saturday, January 07, 2011
UKOUG was UKOUG ... the world's best user group conference. There were a few rough edges but ...
enough to make UKOUG leadership a bit uncomfortable but not enough to detract from the overall experience which was the reason why I come every year.
This year was a bit different for me. I arrived without a speaking spot ... something that thanks to bad weather and good friends didn't last for long.
Within hours of the conference opening I had two spots and presented on Edition Based Redefinition and Developing Extremely large databases.
We seem to be at a point in our industry where we need to redefine some terminology so I will attempt to do so here with a modest proposal.
Very Large Database
1- 100 TB
Extremely Large Database
100 - 250TB
Amazingly Large Database
250 - 500TB
Staggeringly Large Database
500 - 1000TB
Insanely Large Database
I've no doubt, in a few years, this will become laughable as the monsters we manage keep getting larger and larger.