Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


AX-61 NVME issues?
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.

AX-61 NVME issues?

We use the AX61 NVME with CL8 and cpanel, its an upgraded server from older model from hetzner,
but this server facing some extremely weird ISSUES,
random freeze with no outputs, that require hard physical reboot as for hetzner saying nothing shows at screen, and when cold boot is done, server boot with not errors.

Nothing shows on logs at the server, CL support tried to see also for few weeks debugging it with no success for what causing the issue.

When server crash it doesnt under any heavy load, could happen also with 7% cpu usage or 19%, its just random could be stable for a week, or two weeks, or few days and then sudden freeze.

now i wonder if it related to CPU compabillity with CL? as i seen in old topic using AMD is hit and miss with cloudlinux for stabillity?
cause i have exact same server that stable for long time..

should i go with enterprise grade cpu of intel?
like -
https://www.hetzner.com/dedicated-rootserver/px93
or maybe -
https://www.hetzner.com/dedicated-rootserver/ax161

or go for another AX-61 NVME? and maybe some faulty hardware invovled like faulty RAM?

I'm pretty lost and would like to get people advise on it?

Comments

  • What OS ? KErnel ? Try with Archlinux and latest kernel(s) and see if issue is same. I had something similar on AX101 (with Arch) and had to add couple parameters to grub cmd line

  • @amarc said:
    What OS ? KErnel ? Try with Archlinux and latest kernel(s) and see if issue is same. I had something similar on AX101 (with Arch) and had to add couple parameters to grub cmd line

    Hey amarcc, thanks for replying,
    OS is Cloudlinux 8 kernel version is -
    4.18.0-305.17.1.lve.el8.x86_64
    i cant try Archlinux this server using cPanel >:

  • CConnerCConner Member, Host Rep

    Had a same-ish issue on a bunch of Hetzner's 9900k machines. Not sure what the issue was in the end, we just swapped them for a different model and moved on.

  • @Advicerxyz said: Hey amarcc, thanks for replying,

    OS is Cloudlinux 8 kernel version is -
    4.18.0-305.17.1.lve.el8.x86_64
    i cant try Archlinux this server using cPanel >:

    Well somehow you will need latest 5.1x kernel and see if that fixes issue(s)

  • @CConner said:
    Had a same-ish issue on a bunch of Hetzner's 9900k machines. Not sure what the issue was in the end, we just swapped them for a different model and moved on.

    We had the hetzner 9900k for 2 years without issues,
    but seen in reddit also someone facing issue with the AX61

    My thinking now when i order a new server maybe just go for intel now for more stable linux kernel?

  • donkodonko Member
    edited October 2021

    I had this issue but on 2 auction servers (i7 4770) never found the problem i just replaced servers, also currently my ax61 works ok with CL7 hybrid kernel and cpanel

  • This issue exists on ALL Hetzner servers. I got over 150 servers with them that are super stable and run without any issues, however when I order another batch you can sort them out within hours after delivery. If you get any server with them that crash for no reason under minimal load, cancel it directly or ask for a full replacement - this issue can't be fixed by them. Their stress test under Linux reports often no issues at all, which reduces your chance of further support. A workaround which might work for you: undervolt the CPU with the Ryzen Master. You can keep the clock speeds but reduce the Voltage. It should solve your issue, but it's only a workaround for an issue that literally exists across all their datacenters. I would love to know what exactly causes this issue (Motherboard/PSU?) but even after months im still in the dark.

  • PUSHR_VictorPUSHR_Victor Member, Host Rep
    edited October 2021

    I have an AX51 in Finland that does the same. Every now and then it goes down and requires a reboot. Log is completely silent.

    Edit: thats Ubuntu with Nginx on it. No Cloudlinux or anything else.

  • You cannot undervolt the current generation of hetzner Ryzen servers, the Mainboard has an overclock / voltage lock. So Ryzen Master is useless. Don't think you can change that in bios without flashing a modified bios

  • now its drives me mad, i have ordered another server of px93, less than a week sudden freeze, luckly i was able to remote reboot it, but nothing in the logs, im really lost with this one...

  • I've had a few AX-61 servers with Hetzner in the past and have had no issues with them. Like others suggested, I would ask them to replace the hardware or have them transfer you to another server without the problem. I was able to replicate the issue using Ryzen 5950x servers in the United States.

  • @MythHost said:
    I've had a few AX-61 servers with Hetzner in the past and have had no issues with them. Like others suggested, I would ask them to replace the hardware or have them transfer you to another server without the problem. I was able to replicate the issue using Ryzen 5950x servers in the United States.

    I have ordered PX93 and less than a week it frooze again since ordering, nothing in logs,
    the px93 i can reboot in hetzner robot not like the ryzen that needed cold boot,
    its just normal Centos 8.4 minimal(preisntalled from robot), than installed cpanel and converted to Cloudlinux 8.

    The issue if they will replace hardware and remain the hard drives, as far as i know hetzner doesnt replace the same models of hardware, server could refuse to boot up isnt it?

    but also i wonder WHY two servers in a row diffrent model have around the same issues

  • @Cesslor said:
    This issue exists on ALL Hetzner servers. I got over 150 servers with them that are super stable and run without any issues, however when I order another batch you can sort them out within hours after delivery. If you get any server with them that crash for no reason under minimal load, cancel it directly or ask for a full replacement - this issue can't be fixed by them. Their stress test under Linux reports often no issues at all, which reduces your chance of further support. A workaround which might work for you: undervolt the CPU with the Ryzen Master. You can keep the clock speeds but reduce the Voltage. It should solve your issue, but it's only a workaround for an issue that literally exists across all their datacenters. I would love to know what exactly causes this issue (Motherboard/PSU?) but even after months im still in the dark.

    it looks like psu doesnt have enough juice !

  • Hetzner_OLHetzner_OL Member, Top Host

    Hi @Advicerxyz - I asked the hardware team about this. They pointed out that a kernel 5.4+ is recommended for any of our Ryzen 3000 servers and 5.12+ for Ryzen 5000. So keep troubleshooting with that in mind. However, if additional troubleshooting doesn't work, please keep communicating with our support team and documenting what you can. --Katie

  • @Hetzner_OL said:
    Hi @Advicerxyz - I asked the hardware team about this. They pointed out that a kernel 5.4+ is recommended for any of our Ryzen 3000 servers and 5.12+ for Ryzen 5000. So keep troubleshooting with that in mind. However, if additional troubleshooting doesn't work, please keep communicating with our support team and documenting what you can. --Katie

    Hey Katie, we use Cloudlinux as cloudlinux latest kernel is 4.18 based of redhat,
    but it doesnt matter we ordered another server as upgrade number 3 with the server PX93, that also freezes/hang, luckly at least the PX93 not required hetzner to do physical reboot.
    but its frustrating also when the PX93 freeze/hang were at 4 am, and it happened after 6 days of uptime with cpu use of 4%,
    the server not running some massive applications but just cpanel and websites.

Sign In or Register to comment.