The Moment You Discover Your Hardware is Faulty and You Are Not Happy At All
I just love it when I'm building a high-end PC and something just happens to fail and you start looking where exactly the problem is occurring and what is causing the issue, preventing the system to work properly. That Dual Xeon CPU system with 20 cores and 40 logical threads, equipped with 64 Gigabytes of RAM suddenly starts to behave weird as soon as you build it, so you start testing the processors, the motherboard and the memory to see what might be causing it.
Building dual processor systems often can lead to some issues, sometimes they are software related, sometimes they are hardware related... it could be the motherboard BIOS, the two processors just not working well together, the memory controller not liking the RAM modules, or a hardware issue with some of the components. Finally when I have moved the Xeon processors on a single CPU X99 motherboard with 4 different RAM modules to test all four memory channels for each of the processors separately and then the problem suddenly appears...
The problem is one of the channels of the memory controller of one of the CPUs, so I'm finally relieved finding the problem, but it took me a couple of hours until I tested everything and figured where the issue lies. I have started with not all of the memory channels populated and that was a mistake on my side as everything was working just fine with memory on two out of the four memory channels. Fortunately the other processor was just fine, so only one needs to be replaced... that will take some more extra time, not very happy, but just the way things are.
If you have a question or want to add something, then please leave a comment below.
Did you like what you have just read? Check my other posts on steemit @cryptos
If you like what I'm doing for Steem and on Steemit you can support me as a Witness
@cryptos
This is some beast that you are building. Great.
I am glad that you found the cause. It is a game of slowly eliminating the possibilities from the most obvious to the least obvious. Although we tend to skip THE MOST obvious one only to return to it at the end and to find out it is the cause of the error :)
I am much more a user that a builder. I ordered my new machine last week. I put it together from components with most important ones being i7 6700 CPU, GeForce 750 GPU and 2x16 GB of RAM. For office work until WIndows 20 and Ubuntu 32, I hope :)
If it is only for office work it might actually do until then with these specs :)
:) Yes, i has to do. I am buying my desktops to last for some years. At least I hope so.
Anyways, it will do for office work and several VMs for running various alpha and beta nodes of Steemit, Synereo and whatever else comes in the future...
BTW: I applied for the curator job :) A question - do curators generally vote on a posts less that 24 hours young? It seems logical to me so that your bot can catch the vote up in time...
Most of the posts that get votes are new ones, still in their first payput period.
OK, I thought so.
Welcome to high end PC modding!
I've been doing that for many years already... :)
Working with Xeon's I kinda guessed that :) GOOD LUCK MAN!
This post has been linked to from another place on Steem.
Learn more about linkback bot v0.4. Upvote if you want the bot to continue posting linkbacks for your posts. Flag if otherwise.
Built by @ontofractal