Difference between revisions of "SMEServer v8.0 in OpenVZ"
Line 103: | Line 103: | ||
* YUM install sme rpms | * YUM install sme rpms | ||
yum --disablerepo=base,updates,extras upgrade --nogpg | yum --disablerepo=base,updates,extras upgrade --nogpg | ||
− | yum install smeserver\* e-smith\* --enablerepo=smeos8,smeupdates8 --nogpgcheck | + | yum install smeserver\* e-smith\* *.sme.* --enablerepo=smeos8,smeupdates8 --nogpgcheck |
* 287 MB / 374 RPMs Installed | * 287 MB / 374 RPMs Installed | ||
* SMEServer uses runlevels 7 and 4 | * SMEServer uses runlevels 7 and 4 |
Revision as of 10:43, 1 March 2013
SMEServer v8.0 in OpenVZ
This is a work in progress and does not yet work.
Initial CentOS Virtual Machine
- Start with a CentOS 5 Minimal Template in Proxmox VE OpenVZ Container
- Rename template to conform to ProxmoxVE naming convention and move to template cache
cd /var/lib/vz/template/cache wget http://download.openvz.org/contrib/template/precreated/centos-5-i386-minimal.tar.gz mv centos-5-i386-minimal.tar.gz centos-5-minimal_5_i386.tar.gz
- 512 MB RAM / 8 GB HDD / IP address assigned during VM creation / SELinux disabled / sshd enabled
- Start Container
- 88 RPMs pre-installed list
rpm -qa --qf "%{n}-%{v}-%{r}.%{arch}.rpm\n" > /tmp/centOS5_minimal_after_yum_with_arch.txt
- 590 RPMs in SME8 installed list - on Atom Hardware with 2GB RAM / 320 GB HDD
- Check the current CentOS Version with cat /etc/redhat-release to be CentOS release 5 (Final)
Install wget
rpm -Uvh http://vault.centos.org/5.8/os/i386/CentOS/wget-1.11.4-2.el5_4.1.i386.rpm
Install some certificates
rpm --import http://sme-mirror.firewall-services.com/releases/8/smeos/i386/RPM-GPG-KEY-SMEServer rpm --import http://apt.sw.be/RPM-GPG-KEY.dag.txt rpm --import https://fedoraproject.org/static/217521F6.txt # rpm --import http://mirror.centos.org/centos/RPM-GPG-KEY-CentOS-5 rpm --import http://www.salstar.sk/pub/yum/keys/RPM-GPG-KEY-Fedora-Pre-Extras # rpm --import http://dag.wieers.com/packages/RPM-GPG-KEY.dag.txt # rpm --import http://dl.fedoraproject.org/pub/epel/RPM-GPG-KEY-EPEL
https error workaround
[root@cent ~]# rpm --import https://fedoraproject.org/static/217521F6.txt error: https://fedoraproject.org/static/217521F6.txt: import read failed(-1). wget https://fedoraproject.org/static/217521F6.txt --no-check-certificate rpm --import 217521F6.txt rm -f 217521F6.txt
Install Yum Package Manager
rpm -Uvh \ http://vault.centos.org/5.8/os/i386/CentOS/elfutils-0.137-3.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/elfutils-libelf-0.137-3.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/elfutils-libs-0.137-3.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/expat-1.95.8-8.3.el5_5.3.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/gmp-4.1.4-10.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/m2crypto-0.16-8.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/readline-5.1-3.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/sqlite-3.3.6-5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/libxml2-2.6.26-2.1.12.el5_7.2.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/python-2.4.3-46.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/libxml2-python-2.6.26-2.1.12.el5_7.2.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/python-libs-2.4.3-46.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/python-elementtree-1.2.6-5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/python-iniparse-0.2.3-4.el5.noarch.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/python-sqlite-1.1.7-1.2.1.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/python-urlgrabber-3.1.0-6.el5.noarch.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/logrotate-3.7.4-12.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/nss-3.12.10-8.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/nspr-4.8.8-2.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/popt-1.10.2.3-27.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/rpm-4.4.2.3-27.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/rpm-libs-4.4.2.3-27.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/rpm-python-4.4.2.3-27.el5.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/yum-3.2.22-39.el5.centos.noarch.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/yum-fastestmirror-1.1.16-21.el5.centos.noarch.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/yum-metadata-parser-1.1.2-3.el5.centos.i386.rpm \ http://vault.centos.org/5.8/os/i386/CentOS/yum-protect-packages-1.1.16-21.el5.centos.noarch.rpm
- Check the installed groups with yum grouplist:
Installed Groups: System Tools Yum Utilities Available Groups: Administration ....
The RPMs that now stand installed are listed here..
Install screen
To enable background updates, screen is installed.
yum install screen nano mlocate
It is invoked / re-attached with screen -aR scrname and detached from the session with Ctrl-A d.
Set keepcache=1 in /etc/yum.conf to retain the rpms for further use.
Take a backup of the Virtual Machine now before the mess that begins next.
Install the SME 8 RPMs
- Remove all current repos in /etc/yum.repos.d
mkdir /tmp/repo.bak mv /etc/yum.repos.d/*.repo /tmp/repo.bak/
- Upload /etc/yum.repos.d/sme_base.repo
wget http://wiki.contribs.org/images/3/30/SME8_sme-base.repo.txt -O /etc/yum.repos.d/sme-base.repo
- YUM install sme rpms
yum --disablerepo=base,updates,extras upgrade --nogpg yum install smeserver\* e-smith\* *.sme.* --enablerepo=smeos8,smeupdates8 --nogpgcheck
- 287 MB / 374 RPMs Installed
- SMEServer uses runlevels 7 and 4
rm -f /etc/rc4.d/* cp -a /etc/rc7.d/* /etc/rc4.d/
- Now get the list of installed RPMs
ABC=`rpm -qa` echo ${ABC} > sme8_rpms_after_yum.txt ABC=
- Some RPMs get removed and others get updated.
- Use a MySQL database to check the differences:
CREATE DATABASE `rpmmatch`CHARACTER SET latin1 COLLATE latin1_general_ci;
CREATE TABLE `rpmmatch`.`smerpms`( `rpmname` VARCHAR(100) NOT NULL, PRIMARY KEY (`rpmname`) );
CREATE TABLE `rpmmatch`.`c5minrpms`( `rpmname` VARCHAR(100) NOT NULL, PRIMARY KEY (`rpmname`) );
LOAD DATA LOW_PRIORITY LOCAL INFILE 'C:\\smerpms.csv' INTO TABLE `rpmmatch`.`smerpms` FIELDS ESCAPED BY '\\' TERMINATED BY '\t' OPTIONALLY ENCLOSED BY '"' LINES TERMINATED BY '\r\n' (`rpmname`);
LOAD DATA LOW_PRIORITY LOCAL INFILE 'C:\\c5minrpms.csv' INTO TABLE `rpmmatch`.`c5minrpms` FIELDS ESCAPED BY '\\' TERMINATED BY '\t' OPTIONALLY ENCLOSED BY '"' LINES TERMINATED BY '\r\n' (`rpmname`);
UPDATE c5minrpms SET rpmname=REPLACE(rpmname, ".(none).rpm", "") WHERE 1;
UPDATE smerpms SET rpmname=REPLACE(rpmname, ".(none).rpm", "") WHERE 1;
SELECT a.* FROM c5minrpms a LEFT JOIN smerpms b USING (rpmname) WHERE b.rpmname IS NULL ORDER BY rpmname;
SELECT a.* FROM smerpms a LEFT JOIN c5minrpms b USING (rpmname) WHERE b.rpmname IS NULL ORDER BY rpmname;
- 75 packages difference including 2 kernels not needed
Install Missing Packages
Since networking and sshd will not be working, ssh into the ProxmoxVE host and use vzctl enter <VEID> to work further. Edit the file /etc/sysconfig/network-scripts/ifcfg-venet0:
DEVICE=venet0 BOOTPROTO=static ONBOOT=yes IPADDR=127.0.0.1 NETMASK=255.255.255.255 BROADCAST=0.0.0.0 IPV6INIT="yes"
to be something like:
DEVICE=venet0 BOOTPROTO=static ONBOOT=yes IPADDR=192.168.60.110 NETMASK=255.255.255.0 GATEWAY=192.168.60.1 NETWORK=192.168.60.0 BROADCAST=192.168.60.255 IPV6INIT="yes"
Restart Networking with:
ifdown venet0 && ifup venet0
Alternatively, the VLANs can be used:
# cat /etc/sysconfig/network-scripts/ifcfg-venet0 DEVICE=venet0 BOOTPROTO=static ONBOOT=yes IPADDR=127.0.0.1 NETMASK=255.255.255.255 BROADCAST=0.0.0.0 IPV6INIT="yes" # cat /etc/sysconfig/network-scripts/ifcfg-venet0:0 DEVICE=venet0:0 ONBOOT=yes IPADDR=192.168.22.107 NETMASK=255.255.255.255 ifdown venet0 && ifup venet0 ifdown venet0:0 && ifup venet0:0
Now install the missing packages:
yum install mlocate xz xz-libs \ zip unzip which vixie-cron vim-enhanced vim-common usbutils traceroute tnef time telnet tcpdump \ syslinux sysfsutils strace setserial safecat rsync rp-pppoe rpm-build qmail-qfilter \ pyxf86config perl-suidperl kudzu libgomp lockdev lsof maildrop man minicom mkbootdisk mod_ssl \ mt-st mutt newt anacron at authconfig bc bind-libs bind-utils cdrecord cyrus-sasl \ cyrus-sasl-md5 dhclient dhcp diald diald-top distcache dosfstools eject fetchmail ftp \ gettext glib groff hdparm hesiod iptraf iptstate isdn4k-utils --nogpgcheck
- maildrop complained of key requirement and hence the --nogpgcheck
- pcmciautils is dependent on the kernel and hence avoided
- Restart the Virtual machine and set the IP again as done earlier (we need to address the VM startup sauce to do this atutmagically):
signal-event post-upgrade; signal-event reboot
Package Differences
One RPM from the pre-installed RPMs in the template remains:
beecrypt-4.1.2-10.1.1.i386.rpm
Two kernels and one kernel dependent RPM were not installed:
kernel-2.6.18-308.4.1.el5.i686.rpm kernel-2.6.18-348.1.1.el5.i686.rpm pcmciautils-014-5.i386.rpm
The final list of 574 RPMs and 4 extra keys are here.
Setting locale
/usr/bin/locale
LANG= LC_CTYPE="POSIX" LC_NUMERIC="POSIX" LC_TIME="POSIX" LC_COLLATE="POSIX" LC_MONETARY="POSIX" LC_MESSAGES="POSIX" LC_PAPER="POSIX" LC_NAME="POSIX" LC_ADDRESS="POSIX" LC_TELEPHONE="POSIX" LC_MEASUREMENT="POSIX" LC_IDENTIFICATION="POSIX" LC_ALL=
Change the locale
/usr/bin/localedef -i en_US -f UTF-8 en_US.UTF-8 export LC_ALL=en_US.UTF-8 export LANG=en_US.UTF-8 /usr/bin/locale
The new locale is:
LANG=en_US.UTF-8 LC_CTYPE="en_US.UTF-8" LC_NUMERIC="en_US.UTF-8" LC_TIME="en_US.UTF-8" LC_COLLATE="en_US.UTF-8" LC_MONETARY="en_US.UTF-8" LC_MESSAGES="en_US.UTF-8" LC_PAPER="en_US.UTF-8" LC_NAME="en_US.UTF-8" LC_ADDRESS="en_US.UTF-8" LC_TELEPHONE="en_US.UTF-8" LC_MEASUREMENT="en_US.UTF-8" LC_IDENTIFICATION="en_US.UTF-8" LC_ALL=en_US.UTF-8
Current State of files
df
Filesystem 1K-blocks Used Available Use% Mounted on /dev/simfs 8388608 1344356 7044252 17% / none 549288 4 549284 1% /dev
cat /etc/fstab
#------------------------------------------------------------ # BE CAREFUL WHEN MODIFYING THIS FILE! It is updated automatically # by the SME server software. A few entries are updated during # the template processing of the file and white space is removed, # but otherwise changes to the file are preserved. # For more information, see http://www.e-smith.org/custom/ and # the template fragments in /etc/e-smith/templates/etc/fstab/. # # copyright (C) 2002 Mitel Networks Corporation #------------------------------------------------------------ none /dev/pts devpts rw 0 0
cat /etc/mtab
/dev/simfs / simfs rw,relatime 0 0 sysfs /sys sysfs rw,relatime 0 0 /proc /proc proc rw,relatime 0 0 none /dev tmpfs rw,relatime,mode=755 0 0 none /dev/pts devpts rw,relatime,mode=600,ptmxmode=000 0 0 none /proc/sys/fs/binfmt_misc binfmt_misc rw,relatime 0 0
/var/lib/modules
This folder is empty as no CentOS kernel was installed.
dhcpd
dhcpd was exiting and respawning itself unable to get an IP. IT hardcoded in OpenVZ on NIC side for server only mode. Shut it down with:
service dhcpd stop chkconfig dhcpd off
Errors
Init Boot Log
On startup, the VM emits the following error:
Applying Intel CPU microcode update: FATAL: Could not load /lib/modules/2.6.32-7-pve/modules.dep: No such file or directory /etc/rc.d/init.d/microcode_ctl: microcode device /dev/cpu/microcode doesn't exist? Starting system logger: [ OK ]
The said package is:
rpm -qa | grep -i microcode microcode_ctl-1.17-3.el5
- Removing it would mean the removal of 41 smeserver and e-smith rpms which are dependent.
- Ref: Forum Post.
- Ref: CentOS Docs
- Hence:
/sbin/service microcode_ctl stop /sbin/chkconfig --del microcode_ctl
- Copy the /lib/modules/2.6.32-7-pve folder from the host to the guest.
- Initial Boot Log of the VM will show the following error:
iscsid is stopped Starting iSCSI daemon: WARNING: Error inserting libiscsi (/lib/modules/2.6.32-7-pve/kernel/drivers/scsi/libiscsi.ko): Unknown symbol in module, or unknown parameter (see dmesg) WARNING: Error inserting libiscsi_tcp (/lib/modules/2.6.32-7-pve/kernel/drivers/scsi/libiscsi_tcp.ko): Unknown symbol in module, or unknown parameter (see dmesg) WARNING: Error inserting mdio (/lib/modules/2.6.32-7-pve/kernel/drivers/net/mdio.ko): Unknown symbol in module, or unknown parameter (see dmesg) WARNING: Error inserting cxgb3 (/lib/modules/2.6.32-7-pve/kernel/drivers/net/cxgb3/cxgb3.ko): Unknown symbol in module, or unknown parameter (see dmesg) WARNING: Error inserting libcxgbi (/lib/modules/2.6.32-7-pve/kernel/drivers/scsi/cxgbi/libcxgbi.ko): Unknown symbol in module, or unknown parameter (see dmesg) FATAL: Error inserting cxgb3i (/lib/modules/2.6.32-7-pve/kernel/drivers/scsi/cxgbi/cxgb3i/cxgb3i.ko): Unknown symbol in module, or unknown parameter (see dmesg)
- Remove offending RPM and it's dependancies:
# yum remove iscsi-initiator-utils Removed: iscsi-initiator-utils.i386 0:6.2.0.872-16.el5 Dependency Removed: mkbootdisk.i386 0:1.5.3-2.1 mkinitrd.i386 0:5.1.19.6-79.el5
Remove needless services
Ref: http://wiki.openvz.org/Physical_to_container
rpm --nodeps -e acpid auditd kudzu netplugd microcode_ctl lvm2_monitor irqbalance ntpd lm_sensors
If it complains with some being not installed like auditd netplugd lvm2_monitor ntpd, then
rpm --nodeps -e acpid kudzu microcode_ctl irqbalance lm_sensors
Current Status
- Networking and sshd works
- Pinging outside works
- yum update works
- venet0 and venet0:0 not recognised inside SMEServer
- login as root and type console and enter the configuration menu
- Can test internet access
- elinks does not browse server-manager
- configuration of server fails since no network card is recognised as it relies on udev
- devnodes /dev/tun is available in the container