Really odd behavior, I thought it was hardware at first...

[dbhosttexas]

Okay so here's the scoop. I have 2 identically configured systems. I was planning on a cluster but hadn't got there yet. The host I had hooked up started acting funny, let's lay out the hardware first...

Mind you, these 2 hosts are 100% hardware identical. So if I had a hardware failure on one, the other wouldn't copy the behavior...

AMD Quad Core Athlon 64. Specific model evades me at the moment, will post that up later, doubt it is the issue though....

Mainboard. MSI-907A-G43

Memory. 32GB (4x8) DDRAM (Corsair CML32GX3M4A1600C10 Vengeance LP 32GB (4x8GB) DDR3 1600 MHZ (PC3 12800) Desktop Memory 1.5V)

Hard Drives. Western Digital 1TB SATA II 7200rpm

Optical Drive. Samsung DVDRW SATA.

Video MSI R6450-MD2GD3 Radeon 6450 2GB PCIE.

Add in secondary NIC. InfoZone Realtek dual port PCI-E gigabit lan card.

OS. CentOS 7.x install as Server with GUI with Virtualization Hypervisor.

So what is this thing doing?

Up until last week, it was acting okay, except some odd hangs when loading up VMs.

Last week the host I was using started giving me a snowy screen, and during boot, up until login, the CLI screen was white with black text. Once the GNOME login presented, it would allow me to log in, but the screen would slowly flash, and never brought up the first run screens for gnome.

At frist I thought it was a video board problem. So I swapped out a board with a spare I had. No joy. Okay so maybe a motherboard problem, and I am being lazy, so swap the whole box. Did the OS load on the second box, so I have a fresh, brand new box loaded, and it is doing the same exact thing.

I am stuck and unsure how to proceed with this. DId I manage to hit some bug or did I set it up wrong?

[TrevorH]

Sounds a bit like the last issue in the list of "known issues" in the CentOS 7.6 release notes https://wiki.centos.org/Manuals/Release ... ntOS7.1810

[dbhosttexas]

Well, yep. That kind of explains it doesn't it?

[desertcat]

Okay so here's the scoop. I have 2 identically configured systems. I was planning on a cluster but hadn't got there yet. The host I had hooked up started acting funny, let's lay out the hardware first...

Mind you, these 2 hosts are 100% hardware identical. So if I had a hardware failure on one, the other wouldn't copy the behavior...

AMD Quad Core Athlon 64. Specific model evades me at the moment, will post that up later, doubt it is the issue though....

Mainboard. MSI-907A-G43

Memory. 32GB (4x8) DDRAM (Corsair CML32GX3M4A1600C10 Vengeance LP 32GB (4x8GB) DDR3 1600 MHZ (PC3 12800) Desktop Memory 1.5V)

Hard Drives. Western Digital 1TB SATA II 7200rpm

Optical Drive. Samsung DVDRW SATA.

Video MSI R6450-MD2GD3 Radeon 6450 2GB PCIE.

Add in secondary NIC. InfoZone Realtek dual port PCI-E gigabit lan card.

OS. CentOS 7.x install as Server with GUI with Virtualization Hypervisor.

So what is this thing doing?

Up until last week, it was acting okay, except some odd hangs when loading up VMs.

Last week the host I was using started giving me a snowy screen, and during boot, up until login, the CLI screen was white with black text. Once the GNOME login presented, it would allow me to log in, but the screen would slowly flash, and never brought up the first run screens for gnome.

At frist I thought it was a video board problem. So I swapped out a board with a spare I had. No joy. Okay so maybe a motherboard problem, and I am being lazy, so swap the whole box. Did the OS load on the second box, so I have a fresh, brand new box loaded, and it is doing the same exact thing.

I am stuck and unsure how to proceed with this. DId I manage to hit some bug or did I set it up wrong?

A quick note: Check your SATA Cables and maybe swap them out.

After unintentionally blowing up my workstation by rolling over my machine to 7.6 only to discover a bug that I still can't explain the WHY for, I rolled my machine back to 7.5. Trying to eliminate possible cause I INTENTIONALLY blew it up a second time and went to roll it back... only problem my 2 TB HDD on which the backups are store became rather flaky, and it appeared as though the drive was on its way out the door.... and so we did a whole lot of tests on the drive. The first thing we did was try to copy the backup file off to safe location. After that was done I now had my backup file copied, my buddy wanted me to check all the cable connections, which required me to take the computer down, disconnect it, etc, and root around with flash-light in hand. As long as I had the computer down I decided to swap out the SATA cables for some new ones I had on hand, as occasionally they will go bad for some unknown reason. I changed out the cables and I have not had any problems with my backup drive since.

What I can tell you I was sweating bullets when I thought I had lost my backup for 7.5. SATA cables are cheap -- from < $1.00 -- ~ $5.00. Most of mine are in the $1.50 -- $2.50 range. If your drives are any thing that uses a SATA connection is suspect, suspect the SATA cable and swap it out. You can ask a hardware geek what causes a SATA cable to suddenly go bad. This is at least my 3rd or 4th one that has gone bad on me.

BTW the idea of have to exact mirror image computers is a great idea. My buddy and I have a pair of mirror image computers built at the exact same time, with the exact same parts. If one of us has a problem we can have the other try to replicate the problem, if it can then it is not a hardware problem, if it is not we can quickly narrow down the cause to a few usual suspects. Just for fun IF you have them loaded try a different DE such as KDE or MATE and see if the problem still exists, it would also be interesting to log in as a different user/s as it could be some configuration problem. If you are using GNOME try logging into GNOME Classic.

Hope you solve your problem soon. Have a Merry Christmas and a Safe and Happy New Year.

[ywfn]

I kind of wonder if you really read his post. What do SATA cables have to do with video problems?

[desertcat]

I kind of wonder if you really read his post. What do SATA cables have to do with video problems?

Truthfully probably ZERO... unless you happen store images on a HDD, your video card also connects to a SATA port, etc. You noted that "some odd hangs when loading up VMs", and where are your VM stored?? On a HHD or SSD! ... and where do they connect to?? Yep, your SATA ports. Is it *possible* that these hang ups have something to do with disk accessing your VM that are stored on your HDD/SSD? Possible, but I doubt it, since you are experiencing the EXACT same problem on the computer's mirror image. It is still worth remembering for future reference that one should not rule out the fact that SATA cables do go bad.

Now here is the glory of having two IDENTICAL systems: If Computer 1 is working then something goes wrong, and Computer 2 is working and not having any problems when Computer 1 goes down IF you now switch Computer 2 for Computer 1, and now Computer 2 is having the exact same problem as Computer 1 was having, then the problem may lie outside the computer, say an input cable. Try putting Computer 1 where Computer 2 was located and see if you still have the exact same problems. If so the problem is probably software related -- ask me how I know!!! My workstation *was* working fine until I rolled -the machine over to 7.6. I spent the better part of a week trying to track down the problem, and the problem exists ONLY when running KDE, but not in GNOME, and ONLY when running KDE in 7.6 but NOT in 7.5

You have now successfully replicated the problem on two different machines that are mirror image machines. Either your input is to blame, or this is a software problem. You said you run GNOME. Try a different DE -- say MATE, KDE, etc. Does the problem still exist?? If yes suspect a input problem. Proof: Disconnect the computer from the Network and start it up. If it starts up just fine you have narrowed the search area wherein the problem exists.

Tracking down bugs and stomping on them is a PITA -- been there and done that. But just wait until you finally reach your Great "Ah HA!!" moment. Trust me I do not envy you, and seriously hope you find out what the problem is and how you fixed it. Try not to pull your hair out, as this can be a maddening problem. In the mean time do have a Happy Holiday Season!!

P.S. It is possible that the reason your screen is "snowy" is because... it is SNOWING!!
P.P.S. "Snow" is usually an artefact of electrical interference -- try moving the computer to some other location to see if the problem still occurs.

[dbhosttexas]

Folks, we can close out this issue. It turns out that this is a known bug with the implementation of Radeon support with the 7.6 updates. Kind of a nasty one, but there it is...

One would hope that they are planning on fixing this!