New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Has anyone experienced network outages or freezes with the OVH KS-LE series?
I recently discovered that several of my OVH KS-LE series server were frequently disconnecting from the network.
Unable to ping completely frozen need forced restart.
It is triggered when the network data flow rate is high.
At first, I suspected that their network cards were not cooling properly network card overheating.
I sent a ticket to OVH, but they couldn't find the cause. They said it was fine.
Later...
I found that the problem might be with the 82579V network card with Debian system.
The most troublesome thing is that Proxmox is based on Debian, and I can't leave PVE for the time being, so can't switch to another operating system.
I would like to ask if any other friends have encountered this issue?
How did you resolve it?
Comments
I wouldn't know I just idle them. You actually use them shitboxes?
@allthemtings has several KS-A in production afaik
I have like 9 ks-le.
Mosty ks-le-e
Two ks-le-1
One ks-le-game
Two in EU, rest in BHS.
No issues.
Edit: my network usage is small. Something like 20-30mbps per server on average.
Have you tried disabling offloading and/or disabling aspm and other power management
I've had no issues with my KS-LE
Network no, but freezes, e.g someone at OVH has to give the machine some love until it pings again.
Sorry for your problem.
I can't help, but I have some general experience with OVH hardware issues.
It's no secret they use inexpensive hardware, and while drives can be RAIDed and memory replaced, motherboards are harder to diagnose when random problems appear, and I assume network interfaces are on their motherboards.
I've had roughly 20 servers at OVH in about 13 years. Not all at once, around a handful at a time across different use cases.
At one point, I had 2 identical black-friday Rise servers for a project, in the same datacenter. One of them would occasionally lock up, fall offline, and be rebooted. Sometimes I had to do a hard-boot from the console, and other times it rebooted on its own. When it would drop offline, support would usually move on after seeing it comes up after reboot while ignoring the history.
Since I had two identical hardware running the same software, I could quickly rule out the software as the cause.
Support responded quickly to every ticket, a nice advantage of Rise over Kimsufi and SYS, but the issue was never resolved. Because all their tests passed, they would not move me to a new box, so I ended up cancelling it, which meant the next customer probably inherited the same problem.
If your issue continues and you suspect the NIC, and support is not helpful, migrating to another server may be worth considering. Hopefully your issue is software and can be resolved, but just pointing this out in case you suspect the hardware.
Have done these operations.
I also Disable EEE/TSO/GSO.
These operations are somewhat useful, but not much.
Thanks
..................................