I am currently trying to assist with a couple of long standing Service Requests. This type of work is interesting because the issue is normally pretty deep rooted and complicated. These escalations are also very challenging because lots of very intelligent people have already spent a lot of man-hours investigating the problem.
Anyway, one of the first steps is to review associated Service Requests and try to determine whether they are indeed related to the issue under investigation.
The opening paragraph of one such SR contained this bold assertion from the customer: 'As you can see, we have populated the interface tables correctly and EIM still doesn't work.'
This was a interesting statement and was directly related to the problem I was looking at. This statement went unchallenged by Technical Support so I took it at face value and continued my analysis.
There was further empirical evidence that this data load would fail as changes to the corresponding meta-data in the Siebel Repository had been made in version 7.8. A user key definition on this table had been inactivated which was previously present in 7.5. Interesting.
A few hours later, I returned to this SR as I was about to actually try loading data into the same Siebel tables. This seemed rather pointless if it wasn't going to work. However, I started from scratch, read the documentation and created my own simple test case for a single record.
Sure enough, the customer was correct. Even though, I had populated the correct columns in the correct interface table with the correct data, Siebel failed to populate the target tables.
However, when I reviewed the EIM log files more carefully, the errors were the conventional (foreign key lookup failed) type. I fixed the test data, re-ran the load script and sure enough data appeared in the target tables and was visible in the Siebel application.
So, the customer was actually mistaken and I suspect his bold assertion including the four little words 'As you can see' also misled the technical support engineer.
Last night, my son told me that playing World of Warcraft was much, much quicker since the recent broadband upgrade from 4Mb to 10Mb. This made sense as online gaming presumably needs lots of bandwidth and Virgin customer services told me the upgrade would be effective from last Friday.
Out of curiosity, I downloaded a 100MB file (twice) and looked at the download speeds which were in the range 360-390 KB/sec. Quick but less than you would expect for a 10Mb connection.
I called my friends at Virgin Media to check the status of my broadband upgrade. Sure enough, contrary to what I was told, the 10Mb upgrade will only be activated on 23 March when the engineer installs the V+ box.
So, unless you see it with your own eyes, it didn't happen.
Thank you for your email. I am currently working in a bunker deep underground in the heart of Brussels (near Belgium).
I would normally say 'Please call me directly on my mobile'. However, this secure facility is so secure that no mobile communications are possible.
I would normally pledge to replying to your email on my return. However, I am not quite sure precisely when (or even indeed, if) I ever will return. While I am not literally chained to the desk, the security officer is holding my passport which fills me with a sense of unease and practically equates to the same thing.
The working environment is not ideal. Massive ceiling mounted fans in a server room make for a cool, uncomfortable and noisy environment. Still, this is my punishment for living in an ivory tower far divorced from reality, waving my arms about and drawing architecture diagrams on a whiteboard and I must accept it.
This particular customer insisted that, as a followup to a recent architecture workshop, I return in person to install, configure and test what I foolishly claimed was 'straightforward and trivial'.
As part of my job, I often have the pleasure of walking around call centers where agents use headsets to enable them to interact with a computer while conducting a telephone call with a customer.
I have often wondered whether it would be cool or nerdy to use such a headset myself. To date, where possible, I tend to use my mobile phone on speakerphone so I am free to continue to type. For lengthy conference calls (abroad), this isn't a particularly cost-effective option but please don't tell the CFO.
With the increasing adoption of
VOIP within Oracle and the recent delivery of my deluxe
headset, I am typing this while listening to a team meeting using Cisco
Communicator. This is really neat technology, the sound quality surprisingly good and even better, the telephone call is free, completely free. So please tell the CFO.
As for cool or nerdy, let's just say I will only be doing this from the privacy of my own home or possibly my office.
My kids think my adoption of this technology is absolutely hilarious. They keep smirking whenever they come into the kitchen for a snack and if I happen to be actually speaking on the device, they simply can't contain their hysterics and have to leave the room.
In addition, my wife keeps smiling and repeatedly asking me whether she can get a motor insurance
quote.
Many years ago, a customer thanked me which was a pleasant surprise and quite unusual. Most simply ask 'When will your report be ready ?'
This particular customer remarked on my 'logical and methodical approach' to understanding this long standing complex problem, proposing a plan of action and, on this rare occasion, resolving it.
I was interested by his choice of the words 'logical and methodical'. As an aside, I was convinced I heard a surly DBA murmur 'slow, laboured and hopeless more like' which hurt terribly.
Anyway, I don't believe that I have a 'logical and methodical' approach to problem solving. My approach is born from laziness.
For example, if you have 27 users synchronising 176 objects of 5 different types between two IT systems, there is a lot of data to analyse. When you increase the logging on both systems at all tiers, you get even more data to analyse. So much data that your head explodes.
Reducing the number of synchronising users to 1, the number of objects to be synchronised to 1 and limiting the direction of synchronisation to one-way isn't necessarily logical. It is laziness and merely reduces the amount of log files to a manageable volume so you are merely left with a splitting headache.
Sometimes, it may appear like wasted time and effort to stop staring at a monitor and painstakingly review yet again all the previous history, ask the same questions again and possibly repeat test scenarios that have already been executed many times by miscellaneous disgruntled parties. There is simply no substitute for seeing the evidence with your own eyes.
Similarly, many years ago, I used to quite literally get a detailed trace of a working system and a broken system ('let's look at the good versus the bad') and look for divergence. Occasionally, I still resort to this technique which can still prove to be effective.
This approach can be viewed as a little undergraduate-esque by non-technical people who expect expensive consultants just to waltz in and say 'Oh yeah. I saw this very problem last week in Kazbakistan. You just need to set the undocumented, unsupported and mythical parameter: _firewall_auth=trusted. Right, anyone for lunch ?'
Making use of other resources could also be viewed as lazy but never underestimate the knowledge and experience of the available technical resources who may have different backgrounds and areas of expertise and hence a contrary and valuable view on the underlying problem.
People (myself included) often limit their analysis to the areas they know most about.
Right - anyone fancy a walk to that Copenhagen
sports bar for England versus Spain ?