# Intermittent CPU errors: Thank you, thank you, thank you!

Staff Emeritus
Gold Member
It turns out that I have been fighting just this problem with an older industrial computer [industrial automation stuff]. Due to my lack of familiarity with this particular processor, I couldn’t be sure if I had configuration problems or some other software induced problems. I just observed a couple of faults that cinch the problem. THANK GOD!!! This was starting to get nasty. I am sure Zantra and others can appreciate the frustration and potential cost [to me] of problems like this. Consulting work can get real ugly at times.

It is so rare that a processor makes a genuine mistake, especially the stuff I work with, that it takes an act of congress to be sure that no other cause can be found. Industrial computers are very robust and the operating systems are rock solid. This is a very unusual event. Whewwww!

Last edited:

I can relate. Not just to CPU errors, but computer issues in general. Sometimes I just want to take my all purpose computer repair tool(hammer) and fix it for good

Tsu
Gold Member
Originally posted by Zantra
I can relate. Not just to CPU errors, but computer issues in general. Sometimes I just want to take my all purpose computer repair tool(hammer) and fix it for good
I read a story of police responding to a home after getting a report of 'shots fired'. Upon arrival, a man opened the door holding a 12-gauge shot gun - the officers could see behind him that his computer had been shot to hell and back. He quietly admitted to having committed 'computercide'.

Staff Emeritus
Gold Member
Originally posted by Zantra
In memory of Fruit_loops 1994-2004

As for computer problems and hammers [or shotguns!], ain't it the truth!

I'm not sure how your work differs from mine, but typically, unless I design the system, I am the last person in a long design process that involves many engineers...and even worse, budget managers! All of the design and engineering mistakes, along with any errors in component selection, the operational requirements or installation, and any mistakes made in the systems specifications are manifest when I begin to test my program. It is very difficult for me not to eat time for other people’s mistakes. When you throw equipment failures or flaws in the mix, it can get really, really ugly. Then, when I do finally prove that my work is fine and that other problems exist, no one wants to pay for the time. After 20 or 30 meetings and perhaps hundreds of phone conversations, it is often near to impossible to determine exactly who is liable.

I had a really ugly job once that used Modicon/Telemechanique - a French computer system. I am used to American, German, and Japanese systems, but this one takes the prize. After eating about a month's worth of work, and after paying for about 3 weeks of hotels, it finally came out that the components used had never talked with each other before. The main CPU and the field units did not talk the same language...exactly. I got to be the factory R&D man - in the field [actually 25 feet underground] and at my expense.

Self employment: When it’s good, it’s really good. When it’s bad, it’s really, really, really bad.

I have only seen one other intermittent CPU failure over the last ten years. In this case it, for about three days it appeared that I had personally [mostly] shut down the cement business for the entire northwest sector of the US. I spent one weekend in a state of absolute panic. The failure was costing no less than $10,000 an hour; 24 hours a day, everyday. I didn't understand what was happening, and the Port of Portland was breathing flames down my neck. At the time, I did not have the insurance to cover such an event. , Last edited: Originally posted by Ivan Seeking Did you lose your buddy? As for computer problems and hammers [or shotguns!], ain't it the truth! I'm not sure how your work differs from mine, but typically, unless I design the system, I am the last person in a long design process that involves many engineers...and even worse, budget managers! All of the design and engineering mistakes, along with any errors in component selection, the operational requirements or installation, and any mistakes made in the systems specifications are manifest when I begin to test my program. It is very difficult for me not to eat time for other people’s mistakes. When you throw equipment failures or flaws in the mix, it can get really, really ugly. Then, when I do finally prove that my work is fine and that other problems exist, no one wants to pay for the time. After 20 or 30 meetings and perhaps hundreds of phone conversations, it is often near to impossible to determine exactly who is liable. I had a really ugly job once that used Modicon/Telemechanique - a French computer system. I am used to American, German, and Japanese systems, but this one takes the prize. After eating about a month's worth of work, and after paying for about 3 weeks of hotels, it finally came out that the components used had never talked with each other before. The main CPU and the field units did not talk the same language...exactly. I got to be the factory R&D man - in the field [actually 25 feet underground] and at my expense. Self employment: When it’s good, it’s really good. When it’s bad, it’s really, really, really bad. I have only seen one other intermittent CPU failure over the last ten years. In this case it, for about three days it appeared that I had personally [mostly] shut down the cement business for the entire northwest sector of the US. I spent one weekend in a state of absolute panic. The failure was costing no less than$10,000 an hour; 24 hours a day, everyday. I didn't understand what was happening, and the Port of Portland was breathing flames down my neck. At the time, I did not have the insurance to cover such an event. ,

I think we're in slightly different fields. Mine is more of an operational perspective. I handle the network related functions, and LAN/WAN connectivity.

You build em, I fix em when they break But it's the same principle. I'm sure it's frustrating when the buck stops at you, and I must admit there's something to be said for the safety of coporate culpability. Of course that's only the pro...

Oh and fruit_loops was the young boy who recently passed away from cancer

Staff Emeritus
Gold Member
Originally posted by Zantra
Oh and fruit_loops was the young boy who recently passed away from cancer

Ah. Tsunami was telling me about that thread but I never caught his name.

Oh and our company does have a system that requires 99.9 percent uptime, and it does cost the company $10,000 for every minute it's down. Luckily, that's not my jobd Staff Emeritus Science Advisor Gold Member Originally posted by Zantra$10,000 for every minute it's down.

I don't think I could manage the insurance for that one. Technically though, the only thing specified as beyond the scope of my insurance are Nuclear Power Plants and Nuclear Weapons systems.

I don't think I have ever experienced that kind of liability – \$10,000/minute. Usually, my biggest concerns are catastrophic failures and other sources of injury or death. I remember turning down a job for wire factory that was filled with incredibly dangerous open presses, cutters, and rollers; all real heavy duty, limb removing stuff. I swear this is true: I saw perhaps a dozen people with a missing body part and they have a one armed safety manager named Lefty. When I met Lefty, I left. This was one of the ugliest places I have ever seen.

How do you avoid or cover your liability for professional errors and omissions; or do you operate under your customers insurance?

self edited so I didn't give away any secrets I'm not supposed to.

Last edited: