joeware - never stop exploring... :)

In actuality cheap I.T. labor screws I.T. up and ends up costing more in quality, correction/redo work, and loss of customers sick to death of dealing with your shitty workers. You get what you pay for. If you want to use dirt cheap I.T. labor… when your data and everything else is ass backwards, you know why. When all of your good people are saying, “You are screwed, I’m sick of dealing with the morons you are hiring to do this work that force me to work two, three, or more times harder just to keep things afloat, I’m out of here…” You know why… If you get paid at best $20k USD for a year worth of I.T. work, you do not have that job because you know what you are doing, you have that job because you are cheap and you are the reason for this blog post.

In every case where you think cheap “run book” driven I.T. labor is a good idea it can almost certainly be automated and actually run properly versus being screwed up by the huge number of cheap labor workers you throw at the problem. Even the car industry realized that it was better to use robots than people pretending to be automatons because the robots are less likely to screw up so you get better quality with the robots. The problem here is that there is more up front cost which is more immediately visible than dealing with constant quality issues and the extra costs you pay to correct that year after year… So if you have a company struggling to show cost savings now it looks better at the executive level to do the work half ass but cheaply than do it correctly. That is the only reason I can see that execs can justify this stupid thinking that I can visualize.

Why this rant now? Well as I have mentioned before I see using cheap overseas labor for I.T. as a bad long term direction in general for the US/Canada/UK/Europe/Australia/Japan and other true first world countries but mostly because I am neck deep in correcting the crap work put out by these types of resources…

Just today I get asked to run some validation scripts against IPs listed in several thousands rows of an excel spreadsheet which is extract from a support database which is the authoritative database for someone. This is generally no problem, should take seconds to set up and then let the scripts run for how many hours it takes to contact every machine specified… Only the field labeled IP Address doesn’t appear to be standard… Right next to the field is a field indicating the date that someone validated the machine by logging into it, yet they didn’t take the time to make sure the IP address field was in some way shape or form standard… In this field that should be an IP address I found

• IP addresses with octets separated by periods (YEAH!)
• IP addresses with octets separated by commas (WTF? How stupid are we? Where would that ever be valid?)
• IP addresses with octets separated by spaces (yet again… WTF?)
• Multiple IP addresses separated by pipes (|)
• Multiple IP addresses separated by tildas (~)
• Multiple IP addresses separated by spaces ( )
• Multiple IP addresses separated by commas (,)
• Multiple IP addresses separated by semi-colons (;)
• Multiple IP addresses separated by slashes (/\)
• Multiple IP addresses separated by newline characters (/n)  – This is especially fun to see in Excel…
• Multiple IP addresses separated by random parens use
• Multiple IP addresses separated in some random way but with labels mixed in like Ext, Int, Cl, etc.
• Multiple IP addresses separated by a random combination of all of the above
• Multiple IP addresses separated by one or more of the combinations above but all of the IP addresses are the same
• Multiple IP addresses separated by one or more of the combinations above only not all of the IP addresses are full 4 octet IP addresses
• IP Addresses with unprintable characters embedded like 1.2.3.4áááááááááá
• Random use of quote characters ” and ‘ – This is fun to see in Excel, it covers it up…

The data is just so bad off that trying to automate the cleanup of it is impossible, all I can do is have it report anything that doesn’t match a specific format so I can manually correct the spreadsheet. Had the formatting been locked down right off or proper logging was in place for the updates so you could find and beat the people responsible or you just used intelligent people in the first place this wouldn’t be an issue now…

I won’t even go into the column with the FQDNs other than it is immensely obvious not everyone knows what a Fully Qualified Domain Name is.

I realize that possibly someone wasn’t told they should clean this up but something you can often get out of more intelligent (read more expensive) workers is a thing called common sense… I could not have opened this DB and saw those fields and not commented to someone somewhere that it needed to be corrected the instant I laid my eyes on it. In this case, the database has been in use for literally months and now they realize they have an issue with some of the naming info so want me to validate it only the field you validate it from itself is completely hosed. I wish I could get the names of the people who populated the IP Address field in the various random ways and just sit down with them and ask… Why? Why did you choose to populate that field the way you did? Did you not look at the other fields to see the formatting? Did you not think to ask someone? Does what you did actually make sense to you or do you make so little that you just don’t give a shit and figure any old crap will suffice?

You want to reduce the cost of your I.T…. Hire intelligent people who know how to automate things and pay them enough to give a shit. You want people who are constantly thinking and *want* things to run well. People thinking “how can this be automated and then never require a human to put their hands on it again?” A lot of companies have a lot of computer people who aren’t actually very good at using computers to help them do their jobs. They seem to think that they are there to make the computers run well so that means humans must do manual work versus making the computers take care of themselves and each other.

   joe