Tag Info

New answers tagged

0

Yes, if your provider also provides a separate network for this purpose. Otherwise just use the public IP addresses. Though in this case, if you use SSL you will have to pass it all the way through to your backend servers rather than terminating it at the load balancers, to ensure that other customers' virtual machines cannot sniff your traffic. You'll also ...


0

the page give by Aleksandr Levchuk give the following script CTID=777 echo '#!/bin/bash . /etc/vz/vz.conf . ${VE_CONFFILE} SRC=/mnt/disk DST=/mnt/disk if [ ! -e ${VE_ROOT}${DST} ]; then mkdir -p ${VE_ROOT}${DST}; fi mount -n -t simfs ${SRC} ${VE_ROOT}${DST} -o ${SRC} ' > /etc/vz/conf/${CTID}.mount chmod +x /etc/vz/conf/${CTID}.mount ...


0

This messages looks like a bug, please upgrade to lastest current OpenVZ kernel and userspace, after that recheck this operations. If they fails again, please send dmesg content to http://bugzilla.openvz.org


0

You can try my manual: http://www.stableit.ru/2013/11/how-to-mount-multiple-disks-to-openvz.html If you have an account in OpenVZ bugzilla please add write few words about why this feature required for you: https://bugzilla.openvz.org/show_bug.cgi?id=2754.


0

Are you really sure about LXC support in OpenVZ WebPanel? AFAIK it need only kernel and userspace from OpenVZ project.


0

The accepted answer is incorrect, based on the official answer: https://wiki.openvz.org/Vzkernel-firmware Correct Answer: You need to remove the vzkernel-firmware by running the following: rpm -e vzkernel-firmware And then adding the following to /etc/yum.repos.d/openvz.repo : exclude=vzkernel-firmware


0

In your system vzkernel-firmware is conflicting with kernel-firmware. Because vzkernel packages are coming from outside the standard package repositories such conflict is not unusual. The easiest way to solve this issue assuming that you don't use original CentOS kernel is to remove it. To do that run: yum remove kernel kernel-firmware Before proceeding ...


0

We expected this problem many times when upgrading OpenVZ. IMHO best solution for this issue is: yum remove kernel. It's pretty safe and solve this issue forever. I have no ideas why you need standard kernel on OpenVZ nodes. But please be careful and check many times about "What packages should remove yum" when removing kernel package.


0

I use Ubuntu 10.04.1 LTS running in openvz. Has been running on 1 core 256 mb for some time but could never get java to work with this configuration. Never paid it that much attention until I needed to run java apps on it and decided to look into it. Couldn't get ca-certificates-java to install. It would fail with open-jdk, default-jdk or when running it ...


0

Looks like this bug related with OpenVZ limit on number of threads for VE. Please check column "failcnt" on row "numproc" in file /proc/user_beancounters.


0

I solved same issue on machine with ploop and modern OpenVZ kernel in this way: http://www.stableit.ru/2013/11/how-to-mount-multiple-disks-to-openvz.html


0

As I supposed, this is a limitation of OPENVZ. A jail is NOT a virtualization.


0

AFAIK, we haven't working and stable solution for this task. Bleeding of edge technology for containers live migration are OpenVZ and P.Haul (part of CRIU project): https://github.com/xemul/p.haul As main requirement I will mention vanilla kernel from kernel.org.


0

I can provide more convenient way for this issue: http://www.stableit.ru/2013/11/how-to-mount-multiple-disks-to-openvz.html


0

It's possible using virtual terminals for KVM VM. You need (keep in mind all my examples for Debian 6/7): Run Linux kernel inside KVM VM with kernel param: console=tty0. it can be done editing /etc/default/grub : GRUB_CMDLINE_LINUX_DEFAULT="quiet console=tty0" Add to KVM VM file /etc/inittab this line:T0:23:respawn:/sbin/getty -l /usr/local/bin/autologin ...


1

At first I recommends you to enable vSwap memory model because it's simpler and more reliable than old UBC styled system. With it you can replace this vzctl set 101 --physpages 0:2G --swappages 0:2G --kmemsize unlimited --privvmpages 2000M:2G --vmguarpages 2000M:2G --oomguarpages 2G --save by vzctl set 101 --physpages 2G --swappages 2G --save After this ...


0

Standard way for solving issue like this is this: Find all container processes using command like this: https://gist.github.com/pavel-odintsov/3e7351ba4ceca33b8cc6 Kill all container processes manually using kill -9 If some process can't be killed, check dmesg about hardware errors (in most cases it related with disk or another I/O errors) Get strace -f ...


1

Could you show this command output: cat /proc/mounts from you VPS? There are two disk layout for OpenVZ: 1) simfs (old, obsolet; it's not an block device it's only folder on physical server) 2) ploop (new, recommended; it's block real device with separate filesystem) But on both cases you haven't direct access to block device (like /dev/loopX). But you may ...


0

If you have access to OpenVZ physical node you can create loop device and mount it to VPS. But you can't use loop from OpenVZ VPS because technology restrictions. As another solution you may call your administartor/hosting provider do something like this: http://www.stableit.ru/2013/11/how-to-mount-multiple-disks-to-openvz.html


1

As I recommended you in twitter there are few solutions: vzctl set XXXX --cpumask 0,1 --save And after this operations taskset should work. If it not working yet, please create ticket to bugzilla.openvz.org because it seems as bug.


2

Do you have access to physical OpenVZ node? If yes, you should enable connection tracking using this command: vzctl set XXX --netfilter stateful --save And restart container: vzctl restart XXXX Where XXX is your container ID. If you don't have any access to physical server, you may send this command to your administrator or hosting provider because ...



Top 50 recent answers are included