Finally picked one of these up the other day and was excited to get rolling with it!
Set it up with the latest QuTS Hero OS and the following configuration:
1TB Samsung 980 Pro as the system pool (wanted to play around before my 2 x SN700s arrived and was just going to start fresh to have a mirrored system pool)
16GB Kingston ECC module KSM26SED8/16HD (with a matching stick on the way)
4 x WD Red Plus 12TB drives in RAID 6
PowerShield Defender UPS for a little extra peace of mind
Set this all up with a Ubuntu VM with a USB Blu-Ray drive passed through for media ripping purposes, a Home Assistant VM, a MariaDB instance and finally Emby via the available qpkg.
I started seeding my data in from an external USB hard drive as well as some SMB copies from my old HP N54L Microserver and all seemed to be behaving nicely until the next morning.
The SMB copies failed overnight, so I attempted to log into the QNAP web interface but this wasnāt responding and pinging the IP wasnāt responding. I tried switching the network cable to the other NIC to see if it would pick up another address but no luck here. Hitting the power button seemed to do nothing either so I tried switching the UPS to battery mode to attempt getting it to power down gracefully and nothing after 15 minutes (was set to 2 minute power down) and no luck so I had to resort to hard powering off the unit.
I was hoping that this may have just been a high throughput issue so I changed the Home Assistant VM from 2 cores to 1, powered off the Ubuntu VM and tried continuing the SMB copies but unfortunately this happened 2 more times. Each of these times however, the status LED was flashing red when attempting to power up, but powering off and on resolved that and it powered on fine.
And then it crashed a fourth timeā¦ and would not power on. All I could see was the SBY_LED1 on the motherboard lit red and some activity lights on the NIC but it would not power on no matter what I tried:
Swapping RAM to other slot
Swapping original RAM back in (both slots)
Swapping NVMe to other slot
Pulling all hard drives and NVMe
Pulling RAM (just to see if it would try and fail POST)
Pulling CMOS battery and leaving off for a period and trying to power on again
Powering directly (bypassing UPS)
Alternate powerpoint
Nothing seemed to work until it just randomly powered on a few hours later. Believe me when I say the beep from the unit shocked me!
Thought I was all good until it crashed for a fifth time and it has not come back since despite trying all of the above steps again. Iām pulling my hair out and more than a little disappointed but Iām trying to be patient and waiting for my support ticket to be responded to.
Does anyone have any thoughts or has anyone seen any similar issues in the past??
New units can always fail early on (bathtub curve) I would contact your point of sale first for a swap or refund (that would be faster than a QNAP RMA)
Yeah, definitely one of those things that can happen and just frustrated I canāt play with my new toy haha. I have a call in there too but not sure if theyāll allow it without an RMA approval from QNAP as they only accept emails sadly.
As said, I would have just returned it back to the retailer, no need for an RMA for a brand new unit (unless whatever jurisdiction you are in requires this)
First TS-873A worked a few days before it started āgetting lostā (non-responsive), but would come back with re-boot. With no LCD or vid on the 873A, I was at a lossā¦ Having a TVS-873e that developed a power-up issue just after warranty over, I returned the TS-873A promptly and I expect vendor had seen enough Q doa or quasi-doa that they did not press for supporting doc (qnap support recommended that over slower RMA turnaround).
The TVS-873e I mentioned most would consider dead, but I just try to not power it down, and discovered that on power-up failures, if I would try repeatedly eventually it would work. After getting tired crossing the room to it with each attempt, I saw that I wouldnāt have to wait the 5 minutes for it to fail if I watched the LCD and if āHardware initializationā¦ā display seen within the first minute remained on the screen more than a second or few, that attempt would eventually fail so I hard power-off and try again. Eventually it works, usually within 3-4 tries sometimes 6 or more. Iāve lived with it like that for years. Consider it a āloaded diceā tribute to the love of gambling where unit was born.
A TVS-873 (not sure if any difference from the 873e other than sticker on case) had bays 6-8 die just after warranty died and about 3 years after that bays 3-5 diedā¦ Since then a fully functional āTVS-273ā featuring 6 convenient storage bays w/trays (actually one dead RJ45-GBE (#4) forget when).
Rural Appalachia location service definitely exercises UPSā, but as I said, within days for the 873A before any revert to battery hit. The other failures mentioned were not coincidental with power outages.
Only mentioned those models because x73x family. I recall a TS-853 Pro that worked fine for many years before the forever rebooting loop mentioned in old forum a lot for x53 towers. Also not co-incidental with a power hit. Last year got a TS-1655 and right out of box 2 of the 3 PCIe slots were dead (Be sure to test any functionality you eventually plan to use upon receiving your new device!!!).
I have about a dozen QNAPs now and at least another dozen more fully functional transitioned from in past (to weigh in statistically on your [quote=āToxic, post:8, topic:411ā] strange you have so many fail[/quote] assessment.) Having both tower and rack units, even the low end rack units never seem to have these kind of issues either doa or after years.
Only 2 rack unit issues I can recall. I think the TS-1279U-RP I got long ago was an un-documented ā-R0ā version because its processor was a generation earlier than published specs and only two RAM slots vs. the four published specs cite. When I replaced the two 4GB sticks it came with to two 8GB sticks, dilema: Either stick would work in either slot, but never give me more than 8TB total. I suspect a BIOS issue rather than true hardware failure.
Recalling the only other hardware failure on a rack unit reminds me to comment about support response timeā¦ it seems quick on new unit issues or simple usage questions but much slower on questions on older devices. About 16 years ago 2.5 years into 3 year warranty on a TS-859U-RP started experiencing data corruption. Having no idea what was up, saved notes on observations, re-initialized with different HDDs and restored data. Those were slow and I had a lot of data so that took about a week. Within weeks similar data corruption in a different area. When it happened again within weeks I reached out to support providing detailed doc before/after images showing corruption. After two weeks with the issue support requested that I re-initialize device and restore data. Did so and within weeks similar data corruption. At that point they requested I send them the unit to evaluate. They didnāt respond until after warranty was out. I thought the answer would be āexpiredā warranty, but they said they would RMA it with an upgrade to a re-furbed TS-869U-RP, and 3 months after warranty was expired my NEW TS-869U-RP arrived.
