F5F Stay Refreshed Hardware Desktop Device stopped for over a day and needs to be restarted daily

Device stopped for over a day and needs to be restarted daily

Device stopped for over a day and needs to be restarted daily

Pages (2): 1 2 Next
S
squirelymayo
Member
89
10-20-2022, 04:49 PM
#1
Hey there, I'm working with a Dell Inspiron 5675 and these details: CPU – Ryzen 7 1700X, RAM – 16GB, GPU – RX 6500 XT (previously RTX 3080 FE, now swapped for troubleshooting), Power supply – 650W Seasonic Gold OS – Ubuntu Server 24.04 (64-bit). The main problem is that after a day of use the machine stops working completely: no display output, keyboard/mouse unresponsive, SSH access denied. When the RTX 3080 was present, the RTX GeForce logo would flash when the system crashed. I've tried disabling "Deep Sleep Control" in BIOS but it didn't help. No error messages before the crash. Any suggestions on what might be causing this? Thanks for your help!
S
squirelymayo
10-20-2022, 04:49 PM #1

Hey there, I'm working with a Dell Inspiron 5675 and these details: CPU – Ryzen 7 1700X, RAM – 16GB, GPU – RX 6500 XT (previously RTX 3080 FE, now swapped for troubleshooting), Power supply – 650W Seasonic Gold OS – Ubuntu Server 24.04 (64-bit). The main problem is that after a day of use the machine stops working completely: no display output, keyboard/mouse unresponsive, SSH access denied. When the RTX 3080 was present, the RTX GeForce logo would flash when the system crashed. I've tried disabling "Deep Sleep Control" in BIOS but it didn't help. No error messages before the crash. Any suggestions on what might be causing this? Thanks for your help!

X
xJuanSolo
Junior Member
44
10-22-2022, 11:11 AM
#2
I understand this might seem like a simple check, but does the device function on Windows? This could help pinpoint whether the problem is with software or hardware, guiding you toward the right troubleshooting path. The flashing RXT logo likely depends on software that has stopped working, making it a symptom of a crash rather than the root cause.
X
xJuanSolo
10-22-2022, 11:11 AM #2

I understand this might seem like a simple check, but does the device function on Windows? This could help pinpoint whether the problem is with software or hardware, guiding you toward the right troubleshooting path. The flashing RXT logo likely depends on software that has stopped working, making it a symptom of a crash rather than the root cause.

N
NottaSpy
Member
232
10-23-2022, 04:48 PM
#3
The most probable causes are:
• Bad RAM
• Overheating or unstable CPU
• Kernel power management problems
• Failing SSD/HDD

Check the RAM and run a stress test on the CPU. If the 1700X is outdated, it may be overheating or unstable—keep an eye on temperatures during testing. A crash during the test suggests cooling or power issues. As a last step, update BIOS/kernel or upgrade to a newer mainline kernel if necessary.
N
NottaSpy
10-23-2022, 04:48 PM #3

The most probable causes are:
• Bad RAM
• Overheating or unstable CPU
• Kernel power management problems
• Failing SSD/HDD

Check the RAM and run a stress test on the CPU. If the 1700X is outdated, it may be overheating or unstable—keep an eye on temperatures during testing. A crash during the test suggests cooling or power issues. As a last step, update BIOS/kernel or upgrade to a newer mainline kernel if necessary.

O
owent13
Junior Member
42
10-23-2022, 06:57 PM
#4
It's a solid question. I haven't used the system much recently—about a year or two ago. When I set it up as a home server in December, I swapped the drives and installed Ubuntu Server. I'm planning to test it on my Windows boot disk to see if it behaves the same. Thanks!
O
owent13
10-23-2022, 06:57 PM #4

It's a solid question. I haven't used the system much recently—about a year or two ago. When I set it up as a home server in December, I swapped the drives and installed Ubuntu Server. I'm planning to test it on my Windows boot disk to see if it behaves the same. Thanks!

K
kkatsie
Junior Member
7
10-23-2022, 08:22 PM
#5
Checking RAM performance is possible. There are simpler ways to track temperatures in Linux compared to using bashtop. Since your BIOS hasn’t been updated, you might want to verify it there first. Let me know if you need guidance on that. Thanks!
K
kkatsie
10-23-2022, 08:22 PM #5

Checking RAM performance is possible. There are simpler ways to track temperatures in Linux compared to using bashtop. Since your BIOS hasn’t been updated, you might want to verify it there first. Let me know if you need guidance on that. Thanks!

W
Willz_03
Member
179
10-27-2022, 05:56 PM
#6
What systems are you managing? Can you verify the status after extended operation (like checking CPU usage or temperatures)? Glances provides both resource consumption and temperature data. When using Glances inside a Docker container, it’s reachable through a web browser. It appears the system is gradually consuming resources, possibly due to a memory leak. A 24-hour period is considered long for such issues. Review the system logs for clues about software malfunctions before it becomes unresponsive. Are you operating a Wireguard VPN service using wg-quick? My Ubuntu Server 20.04 LTS tends to slow down after a few hours because of a Wireguard issue that steadily drains system resources. Changing to a Gluetun Docker container resolved the problem.
W
Willz_03
10-27-2022, 05:56 PM #6

What systems are you managing? Can you verify the status after extended operation (like checking CPU usage or temperatures)? Glances provides both resource consumption and temperature data. When using Glances inside a Docker container, it’s reachable through a web browser. It appears the system is gradually consuming resources, possibly due to a memory leak. A 24-hour period is considered long for such issues. Review the system logs for clues about software malfunctions before it becomes unresponsive. Are you operating a Wireguard VPN service using wg-quick? My Ubuntu Server 20.04 LTS tends to slow down after a few hours because of a Wireguard issue that steadily drains system resources. Changing to a Gluetun Docker container resolved the problem.

J
JBRocket
Member
176
11-02-2022, 04:49 PM
#7
Sure, just a clean vanilla Windows installation would help isolate the problem. Let it sit for a while before diving into troubleshooting. Otherwise, you'll waste time with guesswork until you find the real cause.
J
JBRocket
11-02-2022, 04:49 PM #7

Sure, just a clean vanilla Windows installation would help isolate the problem. Let it sit for a while before diving into troubleshooting. Otherwise, you'll waste time with guesswork until you find the real cause.

P
prestoo
Member
65
11-04-2022, 05:07 AM
#8
I recall early EPYC processors often needed a restart after 24 hours due to a freezing issue linked to an uptime timer bug. I don't recall this happening with Ryzen chips, though updating the UEFI BIOS and system settings might resolve it.
P
prestoo
11-04-2022, 05:07 AM #8

I recall early EPYC processors often needed a restart after 24 hours due to a freezing issue linked to an uptime timer bug. I don't recall this happening with Ryzen chips, though updating the UEFI BIOS and system settings might resolve it.

K
KrazyDimi
Member
143
11-05-2022, 04:22 PM
#9
Setting up a few Docker containers on the machine, all originating from Immich. Recent logs show healthy services with timestamps around 12 days ago. One container is running an updated server image, another is initializing PostgreSQL, and a machine learning service is active. No memory issues detected in either Docker or system logs.
K
KrazyDimi
11-05-2022, 04:22 PM #9

Setting up a few Docker containers on the machine, all originating from Immich. Recent logs show healthy services with timestamps around 12 days ago. One container is running an updated server image, another is initializing PostgreSQL, and a machine learning service is active. No memory issues detected in either Docker or system logs.

G
Gosuhunter
Member
57
11-05-2022, 08:30 PM
#10
Completed a full memory test with no issues and stable temperatures around 48-50°C throughout. Updated BIOS from version 1.3.8 to 1.5.0, hoping that resolves the problem. Ran a stress test via stress-ng with high CPU allocation; temperatures peaked at 79°C but never exceeded it. Will let the system rest overnight to check if the BIOS update made a difference.
G
Gosuhunter
11-05-2022, 08:30 PM #10

Completed a full memory test with no issues and stable temperatures around 48-50°C throughout. Updated BIOS from version 1.3.8 to 1.5.0, hoping that resolves the problem. Ran a stress test via stress-ng with high CPU allocation; temperatures peaked at 79°C but never exceeded it. Will let the system rest overnight to check if the BIOS update made a difference.

Pages (2): 1 2 Next