LVE Manager not working correctly

[dan]

I tried out a CloudLinux + cPanel installation on a test server before buying a license, and the whole process went smoothly. I tried an installation on our existing production server, and while the process completed without errors, the LVE Manager is missing a lot of things that it had during my earlier testing:[br][img]http://i.imgur.com/0CEoNmp.png[/img]...mg][br][br]The "Current Usage" tab also remains blank even when there are requests coming in, and while I can get to the page to change package settings, it doesn seem to be behaving in the way it did with the defaults that were set during my test installation, as though LVE isn even being used. What are some things I can look at/test to confirm that things are installed correctly? I am comfortable with a command line, and obviously have root access to the server.

[anup]

I have something similar to report. Its cPanel+CLN 6.x
Statistics tab doesn show all the domains -- even the most busiest sites are missing. It just show 20 odd of 200+ domains.

[ighertesco]

Hello,

Try to re-create lveinfo database:

$ service lvestats stop

$ rm -rf /var/lve/*

$ service lvestats restart

Let us know how it goes.

[dan]

Potential progress: I tried stopping the lvestats service, and it failed. No details were given, but Im assuming its because the service wasnt running. Without removing the contents of /var/lve, I tried starting the service, which also failed:

# service lvestats start

Starting lvestats: Traceback (most recent call last):

File "/usr/sbin/lvestats-server", line 777, in

main()

File "/usr/sbin/lvestats-server", line 739, in main

LVE_VERSION = parseLVEVersion()

File "/usr/lib/python2.6/site-packages/lvehelper.py", line 97, in parseLVEVersion

file = open(/proc/lve/list, r)

IOError: [Errno 2] No such file or directory: /proc/lve/list

[FAILED]

For what its worth, /var/lve/lveinfo.db is empty (0 bytes). Im not sure if thats unusual or not.

[igutsol]

I suppose that you don use CL kernel at current moment on the server. Could you please provide me output of:
# uname -r
# lsmod | grep lve
# ls -l /etc/container/ve.cfg
# grep lvestat /var/log/messages

or you can generate and provide us cldoctorkey to report with all your LVE settings, so we could check and provide you quick solution:
# wget -qq -O - https://www.cloudlinux.com/clinfo/cldoctor.sh|bash
or if there is no wget: # curl -s https://www.cloudlinux.com/clinfo/cldoctor.sh|bash

Waiting for your reply.
Thank you.

[dan]

Looks like youre right. /etc/redhat-release reports "CloudLinux Server release 6.7 (Aleksei Gubarev)", but the kernel is still showing as 4.1.0-x86_64-linode59. The report key I got is b584bcb4-640c-4cc6-b08c-030d3a2f7b4c.

[igutsol]

Hello,

Yes, this is the reason of issue. Did you reboot server after CL installation?
Is this Linode Xen? We have good article on this: http://kb.cloudlinux.com/2014/11/how...nel-on-pv-xen/
Please try install grub:

yum install grub

I see that you have /etc/grub.conf, but file /boot/grub/grub.conf missed, so you need check content of /etc/grub.conf. Then please do steps 5-7 from the article above (or you can provide us outputs of cat /etc/grub.conf and ls -l /etc/grub.conf and well help you).

[dan]

This is a Linode Xen system, and I did reboot the server after installation. /etc/grub.conf is just a dead symlink to /boot/grub/grub.conf.

Linode does also offer a KVM setup, and they claim switching is easy. Would that make things any easier?

[dan]

Also, the article you linked says to remove all non-Xen kernels. I notice that the kernel installed by the kernel-xen package doesn mention xen in its name, as it does in that article:

# yum install kernel-xen

Loaded plugins: fastestmirror, rhnplugin

Setting up Install Process

Loading mirror speeds from cached hostfile

* cloudlinux-x86_64-server-6: xmlrpc.cln.cloudlinux.com

Package kernel-2.6.32-604.16.2.lve1.3.54.el6.x86_64 already installed and latest version

Nothing to do

# rpm -qa|grep kernel

kernel-2.6.32-604.16.2.lve1.3.54.el6.x86_64

dracut-kernel-004-388.el6.noarch

kernel-firmware-2.6.32-604.16.2.lve1.3.54.el6.noarch

kernel-headers-2.6.32-604.16.2.lve1.3.54.el6.x86_64

Is this a problem?

[ighertesco]

Hello,

kernel-xen package is actual for centOS 5 only, CentOS 6 version already contains required modules in kernel package.

It is much easier to setup CloudLinux on Linode KVM: http://docs.cloudlinux.com/cloudlinu...inode_kvm.html, so I would recommend (if it is possible) to order kvm based VM.

[dan]

Sounds good. Ill try that and get back with results.

[dan]

Update: switching the server at Linode over to KVM made all the difference! Once that was done, I re-ran the installer and rebooted, and everything is now working as expected. Thanks for the help.