I have run my NAS and backup server runs on a Supermicro H8DGI-F motherboard for almost 10 years now and have had at least two similar incidents with the motherboard. Once, after working inside of the case to perform regular maintenance, the board failed to boot and would only display a green LED on the motherboard. Another time I had a similar experience but after trying to reset the BIOS with a jumper and rebooting several times I just left it alone on the black screen after turning it on and it "eventually" booted itself up and has continued to run without any problem.

I actually have a box of, I think, two or three replacement boards ready to go, just in case I need them in the future. Considering the cost of parts invested and the time is would take to buy new parts and validate a complete replacement, I decided to spend a couple hundred dollars buying extra motherboards to hedge against any future failures.

I have no interest in replacing this system as I have hundreds of dollars, maybe more, invested in about 256GB of RAM (expandable to 512GB). That was a hell of a lot of memory back in the day and I don't want it have to obsolete it by necessity due to a failure induced upgrade. I will run this system into the ground and it would need to catch fire or suffer some other catastrophic failure before I would even consider replacing it.

I don't think that these boards are unreliable. My guess is that they are just so extremely dense component-wise that there are more opportunities for something to go wrong.

I wanted to highlight though that I too had a seemingly dead server board come back to life. I've always wondered if the first board would have come back to life as well had I left it to sit at bootup like the second one.

I have seen this happen quite a lot. usually reseating something can fix this, as can cleaning the contacts or trying another board or trying at some later day even.

But at some point I just made it a habit to test new items once or twice and if I couldn't get it to work, label it (date and that it failed testing) and put it in a box of all my parts that failed once. Even had a much smaller box of items that failed twice but not sure I still have that since I had to pack everything up some years ago.
But I definitely never throw something out of it failed just once.

I've got an old 440LX micro ATX board that will only boot once every three weeks. Hah. Stupid Trigem quality.

BitWrangler wrote on 2022-02-24, 04:40:

IMO they just think you're a n00b and try to jerk you around, but when you show some resolve, and threaten them with the 150W soldering iron, they cut the crap and play nice.

Same here - or even simpler, I fixed quite a few things, computers or other, by taking them apart and putting them back together.

Thinks, sometimes these machines simply seek some attention.
I assume it takes good karma and you need your chakras in good order and occasionally have to bury a black cat behind your house.

Cuttoon wrote on 2022-04-08, 13:58:

Thinks, sometimes these machines simply seek some attention.
I assume it takes good karma and you need your chakras in good order and occasionally have to bury a black cat behind your house.

It's not attention they are after. It's a thirst for blood. It's well known that pre-atx cases are in fact carnivorous man traps attempting to snare your soft juicy fingers on finely evolved sharp metal edges. Once they get manage to get a little blood, they can generate enough magic smoke to function for a few more years.

douglar wrote on 2022-04-10, 21:34:

Cuttoon wrote on 2022-04-08, 13:58:

Thinks, sometimes these machines simply seek some attention.
I assume it takes good karma and you need your chakras in good order and occasionally have to bury a black cat behind your house.

It's not attention they are after. It's a thirst for blood. It's well known that pre-atx cases are in fact carnivorous man traps attempting to snare your soft juicy fingers on finely evolved sharp metal edges. Once they get manage to get a little blood, they can generate enough magic smoke to function for a few more years.

It's fairly difficult to not get cut, especially if you're careless or tired or impatient, even after you done it for a while. At some point I just started making it a habit to always have some extra bandaids on me whenever I expected to be working inside an AT tower case. It's hard to avoid because the sharp edges can be almost anywhere. It's kinda a funny meme but it's actually kinda true 🤣.
It got better with the advent of ATX at least, though ATX cases could still be just as sharp (especially some of the cheaper crappers out there 😜 ).

I have in the past sliced my hand deep enough on a sharp edge inside of a case that it needed stitches. I still have a scar as a result.

Sort of - was 'fixing' an Intel skt370 cooler fan (noisy bearing) with a little tlc of a stripdown, clean & oil. Set it up to run once done and was now working nice and quiet - 10 minutes later it died!

I have this old Maxtor 5.7GB hard drive that came with the old Acer Aspire 2856 desktop tower PC (since 1998, so I am the original owner). At one point in 2018, it was just stuck, clicking, when powered on. For some reason, I kept the drive. Same thing happened in 2019. Then, in 2020, I tried once again, and it spun up normally and everything. Thoroughly tested the drive and: no bad sectors or anything . I have since imaged the drive that same year, but that was weird.

PC Hoarder Patrol wrote on 2022-04-11, 11:49:

Sort of - was 'fixing' an Intel skt370 cooler fan (noisy bearing) with a little tlc of a stripdown, clean & oil. Set it up to run once done and was now working nice and quiet - 10 minutes later it died!

There isn't much that can go wrong inside of a fan, since most of it is just the electromagnet. It's possible that one of the windings broke. They are very fragile and can server at our near the solder joint with even a small amount of force.

Most old stuff just needs a little coaxing and TLC to work properly again

I had this experience again today with a video card. I had inspected it closely and validated that there was no physical damage. Booted it in a machine that was known to be working and got a black screen and no BIOS. Removed the card, stared at it for a couple of minutes with no problems to be seen, plugged it back in and it worked fine.

Just because it doesn't work the first time doesn't mean there is anything wrong with it, from my experience. And I keep encountering more examples of this.

i stripe a pair of 320gb 2.5in sata drives. every once in a while the same one will fail to spin up and start audibly beeping. i think this means the head is stuck to the platter. power cycling almost never helps and the drive stays in the beeping state UNTIL i power on the drive with the sata cable disconnected then it spins up and brings the array online when i hot plug it. say what you will about the drive being on the brink of failure, but this has been going on for about 6 or 7 years now in my main pc.

Kahenraz wrote on 2022-04-12, 22:14:

I had this experience again today with a video card. I had inspected it closely and validated that there was no physical damage. Booted it in a machine that was known to be working and got a black screen and no BIOS. Removed the card, stared at it for a couple of minutes with no problems to be seen, plugged it back in and it worked fine.

I've had this happen as well. My theory is that, due to improper storage by the previous owner, a thin layer of oxidation had formed on the card's contacts. Or they could just be dirty from dust accumulation and such. Either of these can prevent the card from properly making an electrical connection to the slot on the motherboard. Re-inserting the card a few times might be enough to strip away the dirt/oxidation, so the card appears to magically work at some point.

Thoroughly cleaning the card's contacts with a cotton swab and isopropyl alcohol is often enough to resolve this. But sometimes, gently going over the pins with a soft pencil eraser gets better results. Same thing as with old console cartridges.

stanwebber wrote on 2022-04-13, 02:48:

i stripe a pair of 320gb 2.5in sata drives. every once in a while the same one will fail to spin up and start audibly beeping. i think this means the head is stuck to the platter. power cycling almost never helps and the drive stays in the beeping state UNTIL i power on the drive with the sata cable disconnected then it spins up and brings the array online when i hot plug it. say what you will about the drive being on the brink of failure, but this has been going on for about 6 or 7 years now in my main pc.

I hope you don't keep important data on this drive.

Kahenraz wrote on 2022-04-13, 04:36:

stanwebber wrote on 2022-04-13, 02:48:

i stripe a pair of 320gb 2.5in sata drives. every once in a while the same one will fail to spin up and start audibly beeping. i think this means the head is stuck to the platter. power cycling almost never helps and the drive stays in the beeping state UNTIL i power on the drive with the sata cable disconnected then it spins up and brings the array online when i hot plug it. say what you will about the drive being on the brink of failure, but this has been going on for about 6 or 7 years now in my main pc.

I hope you don't keep important data on this drive.

Important data should ONLY be kept on a RAID 0 array composed of at least 2 of either

- a Quantum Bigfoot
- an IBM Deskstar 75GXP
- a CMI CM6000 series
- a Quantum Plus hardcard
- a Seagate Barracuda 7200.11 with firmware earlier than AD14

otherwise there just isn't enough spice/challenge in life . 😉

If one drive is already constantly teetering on the edge of failure, that's just asking for trouble. If the good drive fails, reconstructing the array will put a lot of stress on the defective drive and could cause it to fail.

I use a RAID10 on my personal workstation. 😀

darry wrote on 2022-04-13, 05:19:

Important data should ONLY be kept on a RAID 0 array composed of at least 2 of either […]
Show full quote

Important data should ONLY be kept on a RAID 0 array composed of at least 2 of either

- a Quantum Bigfoot
- an IBM Deskstar 75GXP
- a CMI CM6000 series
- a Quantum Plus hardcard
- a Seagate Barracuda 7200.11 with firmware earlier than AD14

otherwise there just isn't enough spice/challenge in life . 😉

that's not living dangerously enough for me. i don't just keep live data--i keep backup images of original source code on the drive and it's striped, not mirrored! (this is a joke, but it's also partially true!) after 6 or 7yrs i have no fear now! the powers that be, responsible for this thread, will protect me. any naysaying to the contrary is tantamount to villager screams of BURN THE WITCH to every participant in this thread who also believes in the magic.

I did catch the joke and I liked it. 😀

But not at first. I had to delete and rewrite my original reply, ha ha.

darry wrote on 2022-04-13, 05:19:

- a Seagate Barracuda 7200.11 with firmware earlier than AD14

Hrrrmmm I somehow missed hearing about those, is this all I need to do? http://support.seagate.com/firmware/firmnav_fsc_en.html