OVH OVH dedicated servers

Maybe you have different NVMe units? Mine are the 1.2 TB version. Good question though. What are your server specs?

---

 

I think we're really guessing that it's the NVMe drives. Though, since I'm able to install centos on the non-NVMe box, it would be logical to assume that the NVMe disks are the issue. The box failed again on a netboot with the experimental kernel and another tech opened another ticket. I wrote him back and asked for some more details about the issues. Hopefully, I'll get a tech who knows what's exactly happening.

Here is the specs of the board on the NVMe Box

Code:

Manufacturer: Supermicro
Product Name: X10SRi-F
Version: 1.01B

Here's an additional kicker... was so tired last night, that I ordered the box with 3x 4TB SATA drives. What an effen idiot. Like I'm just throwing money out the door. There goes $167. So pissed right now.

Update: tested Ubuntu 16.10 w/ org kernel on the NVMe box, worked fine. So Ubuntu must have the drivers.

 

Ended up getting the NVMe server working via KVM centos iso install. All the drivers are there and everything is working. Not sure what kernel they're using for their images, but I sent a tweet to the OVH CEO who follows me on Twitter, hopefully he can address the issue.

 

I got: 2x450GB SSD NVMe

 

Honestly, I don't really know what is happening... why it would work on your machine and not mine. My guess would be is you probably have hardware which is supported by the org kernel that OVH is using.

 

But I got different problem first on my new server:
https://community.centminmod.com/threads/have-you-guys-encounter-this-on-ovh-dedicated-server.10675/

OVH tech Support deal with it for several days before I got it installed fine.

 

I haven't gotten to the centmin mod install yet. Not getting the errors you had. Reboots fine. Actually system seems to be working great under the centos iso image. The KVM is really nice. I even got into the bios of the machine and was looking at the settings (had to adjust the boot order to be able to boot from the virtual device which my iso was attached to).

I just hope I get credit for the machine I purchased last night. Otherwise, I've got a $167 boat anchor. Hopefully, they'll give me credit on the existing machine.

 

Code:

hdparm -Tt /dev/nvme0n1

Code:

/dev/nvme0n1:
Timing cached reads:   23194 MB in  2.00 seconds = 11610.94 MB/sec
Timing buffered disk reads: 2660 MB in  3.00 seconds = 886.20 MB/sec
[root@ns541311 mnt]# hdparm -Tt /dev/nvme1n1

Code:

hdparm -Tt /dev/nvme0n1

Code:

/dev/nvme1n1:
Timing cached reads:   23184 MB in  2.00 seconds = 11605.39 MB/sec
Timing buffered disk reads: 1602 MB in  3.00 seconds = 533.56 MB/sec

 

Code:

# modinfo nvme
filename:       /lib/modules/3.10.0-514.el7.x86_64/kernel/drivers/nvme/host/nvme.ko
version:        1.0
license:        GPL
author:         Matthew Wilcox <willy@linux.intel.com>
rhelversion:    7.3
srcversion:     5B1DDA93CDE30B05D22BA64
alias:          pci:v*d*sv*sd*bc01sc08i02*
alias:          pci:v00001C58d00000003sv*sd*bc*sc*i*
alias:          pci:v00008086d00005845sv*sd*bc*sc*i*
alias:          pci:v00008086d00000A54sv*sd*bc*sc*i*
alias:          pci:v00008086d00000A53sv*sd*bc*sc*i*
alias:          pci:v00008086d00000953sv*sd*bc*sc*i*
depends:       
intree:         Y
vermagic:       3.10.0-514.el7.x86_64 SMP mod_unload modversions
signer:         CentOS Linux kernel signing key
sig_key:        D4:88:63:A7:C1:6F:CC:27:41:23:E6:29:8F:74:F0:57:AF:19:FC:54
sig_hashalgo:   sha256
parm:           admin_timeout:timeout in seconds for admin commands (byte)
parm:           io_timeout:timeout in seconds for I/O (byte)
parm:           shutdown_timeout:timeout in seconds for controller shutdown (byte)
parm:           use_threaded_interrupts:int
parm:           use_cmb_sqes:use controller's memory buffer for I/O SQes (bool)
parm:           nvme_major:int
parm:           nvme_char_major:int

Code:

# lsmod | grep nvme
nvme                   61785  8 

Code:

# ls -lahR /lib/modules/$(uname -r)/kernel/drivers/nvme
/lib/modules/3.10.0-514.el7.x86_64/kernel/drivers/nvme:
total 4.0K
drwxr-xr-x.  3 root root   18 Apr  7 14:46 .
drwxr-xr-x. 67 root root 4.0K Apr  7 14:46 ..
drwxr-xr-x.  2 root root   21 Apr  7 14:46 host

/lib/modules/3.10.0-514.el7.x86_64/kernel/drivers/nvme/host:
total 112K
drwxr-xr-x. 2 root root   21 Apr  7 14:46 .
drwxr-xr-x. 3 root root   18 Apr  7 14:46 ..
-rw-r--r--. 1 root root 110K Nov 22 12:29 nvme.ko

Code:

# cat /boot/config-$(uname -r) | grep -i nvme
CONFIG_BLK_DEV_NVME=m

Code:

# uname -r
3.10.0-514.el7.x86_64

 

Interesting that my other server w/ the SATA drives and the OVH centos image (without OVH kernel) has

Code:

# uname -r
3.10.0-514.10.2.el7.x86_64

 

I ran yum update and it doesn't show any additional updates.

Code:

# yum upgrade
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
 * base: mirror.gpmidi.net
 * extras: mirror2.evolution-host.com
 * updates: mirror2.evolution-host.com
No packages marked for update

Code:

# yum -y update kernel
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
 * base: mirror.gpmidi.net
 * extras: mirror2.evolution-host.com
 * updates: mirror2.evolution-host.com
No packages marked for update

 

Figured it out. No kernel-devel. Installed that, now up to date.

Anything else I should install before centmin mod?

 

I figured as much.

I don't know who was using my IP before me, but it's getting hammered by login requests / failures.

 

Code:

# modinfo nvme
filename:       /lib/modules/3.10.0-514.10.2.el7.x86_64/kernel/drivers/nvme/host/nvme.ko
version:        1.0
license:        GPL
author:         Matthew Wilcox <willy@linux.intel.com>
rhelversion:    7.3
srcversion:     80C25968F401246C1527FF3
alias:          pci:v*d*sv*sd*bc01sc08i02*
alias:          pci:v00001C58d00000003sv*sd*bc*sc*i*
alias:          pci:v00008086d00005845sv*sd*bc*sc*i*
alias:          pci:v00008086d00000A54sv*sd*bc*sc*i*
alias:          pci:v00008086d00000A53sv*sd*bc*sc*i*
alias:          pci:v00008086d00000953sv*sd*bc*sc*i*
depends:       
intree:         Y
vermagic:       3.10.0-514.10.2.el7.x86_64 SMP mod_unload modversions
signer:         CentOS Linux kernel signing key
sig_key:        27:F2:04:85:EB:EB:3B:2D:54:AD:D6:1E:57:B3:08:FA:E0:70:F4:1F
sig_hashalgo:   sha256
parm:           admin_timeout:timeout in seconds for admin commands (byte)
parm:           io_timeout:timeout in seconds for I/O (byte)
parm:           shutdown_timeout:timeout in seconds for controller shutdown (byte)
parm:           use_threaded_interrupts:int
parm:           use_cmb_sqes:use controller's memory buffer for I/O SQes (bool)
parm:           nvme_major:int
parm:           nvme_char_major:int

Code:

# lsmod | grep nvme
nvme                   61785  9

Code:

# ls -lahR /lib/modules/$(uname -r)/kernel/drivers/nvme
/lib/modules/3.10.0-514.10.2.el7.x86_64/kernel/drivers/nvme:
total 4.0K
drwxr-xr-x.  3 root root   17 Mar  7 23:48 .
drwxr-xr-x. 67 root root 4.0K Mar  7 23:48 ..
drwxr-xr-x.  2 root root   20 Mar  7 23:48 host

/lib/modules/3.10.0-514.10.2.el7.x86_64/kernel/drivers/nvme/host:
total 112K
drwxr-xr-x. 2 root root   20 Mar  7 23:48 .
drwxr-xr-x. 3 root root   17 Mar  7 23:48 ..
-rw-r--r--. 1 root root 110K Mar  3 08:51 nvme.ko

Code:

# cat /boot/config-$(uname -r) | grep -i nvme
CONFIG_BLK_DEV_NVME=m

Code:

# uname -r
3.10.0-514.10.2.el7.x86_64

 

Not sure what you mean here. Install the OVH image or the Centos ISO? Confirm what works?

NVMe might not be the issue. It might be the mobo. Though, it's odd that the ISO image works and the OVH image doesn't work. Unless they're modifying the Centos org kernel image.

Right now the installer is locked on my machine. Last night I tried to install Ubuntu via the OVH installer and the machine halted on the installer boot. OVH launched an intervention to do a "hardware diagnosis" and the installer is currently locked due to that intervention. It's been 12 hours and I haven't heard back from them. I wonder if I messed up things when I deleted some "Unknown" partitions while installing Centos ISO manually. I would figure that OVH installer would re-create those on a new install or at least they would respond to my ticket and let me know what happened.

Still haven't heard back from the OVH CEO. He was at a conference yesterday. He's pretty good at checking in on issues that are legit. So, we'll see what happens.