升级linux内核(2.6.32->3.10.81)，安装docker

最新推荐文章于 2021-05-28 17:05:27 发布

weixin_30614109

最新推荐文章于 2021-05-28 17:05:27 发布

阅读量217

点赞数

文章标签：运维操作系统开发工具

原文链接：http://www.cnblogs.com/lcword/p/5865562.html

版权

1.内核升级环境准备

#查看已经安装的和未安装的软件包组，来判断我们是否安装了相应的开发环境和开发库； yum grouplist #一般是安装这两个软件包组，这样做会确定你拥有编译时所需的一切工具 yum groupinstall "Development Tools" #你必须这样才能让 make *config 这个指令正确地执行 yum install ncurses-devel #如果你没有 X 环境，这一条可以不用 yum install qt-devel #创建 CentOS-6 内核时需要它们 yum install hmaccalc zlib-devel binutils-devel elfutils-libelf-devel

2.开始升级内核：

 cd /usr/src/linux-3.10.81
 #复制原内核配置 
 cp /boot/config-2.6.32-71.el6.x86_64 .config vim .config /*将如下内容 追加入.config 并保存 CONFIG_NF_NAT_IPV4=y CONFIG_IP_NF_TARGET_MASQUERADE=y CONFIG_NETFILTER_XT_MATCH_ADDRTYPE=y CONFIG_MEMCG_SWAP=y CONFIG_MEMCG_SWAP_ENABLED=y CONFIG_AUFS_FS=y CONFIG_DM_THIN_PROVISIONING=y CONFIG_OVERLAY_FS=y CONFIG_CGROUP_PERF=y CONFIG_CFS_BANDWIDTH=y HYPERVISOR_GUEST=y CONFIG_VMWARE_BALLOON=m */ #编辑内核配置，开启内核CGROUP支持 make menuconfig #选择General setup-->Control Group support->Memory Resource Controller for Control Groups选中 #自动整理.config sh -c 'yes "" | make oldconfig' #编译并安装内核(j8代表8个线程同时编译，请根据你的机器情况设置） make -j8 bzImage && make -j8 modules && make -j8 modules_install && make install vim /etc/grub.conf 修改default=0保存。即选择从你新编译的内核启动linux。 #重启 reboot #注意：重新编译内核请运行 cd /usr/src/linux-3.10.81 make mrproper make clean

3.安装docker

#启动后
uname -r
返回3.10.81表示内核安装并启动成功
#rehl的docker安装说明https://docs.docker.com/installation/rhel/
#如下是安装说明的简化： #a.下载docker的rpm wget https://get.docker.com/rpm/1.7.0/centos-6/RPMS/x86_64/docker-engine-1.7.0-1.el6.x86_64.rpm #本地安装rpm包 sudo yum localinstall --nogpgcheck docker-engine-1.7.0-1.el6.x86_64.rpm #启动docker服务 sudo service docker start 这里等价于docker -d #如果要追加参数请修改/etc/sysconfig/docker 这个文件 如：加入other_args=" -s vfs" -s 是指定使用的文件系统，默认是devicemapper 也可以是vfs 或aufs(ubuntu的） 我的在这里使用默认的devicemapper没有启动成功。请看FAQ# #运行docker的最简镜像 docker run hello-world #docker image/ps/search/pull 等等正常就成功了 #登入容器 docker exec -ti 26723dc2509a /bin/bash

4.FAQ有两个问题还没有解决

ERRO[0000] [graphdriver] prior storage driver "devicemapper" failed: exit status 1
FATA[0000] Error starting daemon: error initializing graphdriver: exit status 1 出现上述错误有如下解决办法！ *1. yum upgrade device-mapper-libs 再次启动看是否可以 2. 使用另一种文件系统试一下如：docker -d -s vfs 默认是devicemapper系统 还可以使用aufs 3.#至此如果还是没有启动docker-daemon守护服务，那就用下面的方法把docker版本降低试下*。 *并在/etc/yum.repos.d/hop5.repo中添加yum源内容* [hop5] name=www.hop5.in Centos Repository baseurl=http://www.hop5.in/yum/el6/ gpgcheck=0 gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-HOP5 保存并更新cache yum makecache 安装docker-io yum install docker-io 再次运行 docker -d 成功！ **上面第3步，启动docker-daemon不成功，可能是我的线上机器是虚拟机的问题所以加上 -s vfs 就启动成功了。** ##宿主机不能正常ping容器的ip 注意宿主（host)的iptables设置。 ##ip netns “Object "netns" is unknown, try "ip help".\n'”报错错误 请安装如下包: wget https://repos.fedorapeople.org/openstack/EOL/openstack-grizzly/epel-6/iproute-2.6.32-130.el6ost.netns.2.x86_64.rpm

5.安装aufs文件系统

cd /etc/yum.repos.d
wget http://www.hop5.in/yum/el6/hop5.repo
yum install kernel-ml-aufs
docker -d -s aufs #使用aufs文件系统启动docker服务

快速安装方法

sudo yum install -y http://vault.centos.org/6.6/os/x86_64/Packages/device-mapper-libs-1.02.90-2.el6.x86_64.rpm http://vault.centos.org/6.6/os/x86_64/Packages/device-mapper-1.02.90-2.el6.x86_64.rpm http://vault.centos.org/6.6/os/x86_64/Packages/device-mapper-event-1.02.90-2.el6.x86_64.rpm http://vault.centos.org/6.6/os/x86_64/Packages/device-mapper-event-libs-1.02.90-2.el6.x86_64.rpm https://dl.fedoraproject.org/pub/epel/testing/6/x86_64/docker-io-1.6.2-1.el6.x86_64.rpm

更多参考：http://dirlt.com/docker.html

示例：
启动容器：docker run -d ubuntu /bin/sh -c “while true; do echo hello world; sleep 2; done”
容器里执行：docker exec ubuntu /bin/sh -c “ls /;”

更简单的升级内核方法，经测试成功！
这里通过yum快速升级CentOS 6.x内核到3.10：
# rpm -ivh http://www.elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm
# yum --enablerepo=elrepo-kernel install kernel-lt -y
在grub.conf中确认装好的内核在哪个位置：
# vi /etc/grub.conf default=0 重启系统，后查看内核信息： # uname -r 3.10.65-1.el6.elrepo.x86_64

检查docker依赖环境的脚本check-config.sh

#!/usr/bin/env bash
set -e

# bits of this were adapted from lxc-checkconfig
# see also https://github.com/lxc/lxc/blob/lxc-1.0.2/src/lxc/lxc-checkconfig.in

possibleConfigs=(
    '/proc/config.gz'
    "/boot/config-$(uname -r)" "/usr/src/linux-$(uname -r)/.config" '/usr/src/linux/.config' ) if [ $# -gt 0 ]; then CONFIG="$1" else : ${CONFIG:="${possibleConfigs[0]}"} fi if ! command -v zgrep &> /dev/null; then zgrep() { zcat "$2" | grep "$1" } fi is_set() { zgrep "CONFIG_$1=[y|m]" "$CONFIG" > /dev/null } # see https://en.wikipedia.org/wiki/ANSI_escape_code#Colors declare -A colors=( [black]=30 [red]=31 [green]=32 [yellow]=33 [blue]=34 [magenta]=35 [cyan]=36 [white]=37 ) color() { color=() if [ "$1" = 'bold' ]; then color+=( '1' ) shift fi if [ $# -gt 0 ] && [ "${colors[$1]}" ]; then color+=( "${colors[$1]}" ) fi local IFS=';' echo -en '\033['"${color[*]}"m } wrap_color() { text="$1" shift color "$@" echo -n "$text" color reset echo } wrap_good() { echo "$(wrap_color "$1" white): $(wrap_color "$2" green)" } wrap_bad() { echo "$(wrap_color "$1" bold): $(wrap_color "$2" bold red)" } wrap_warning() { wrap_color >&2 "$*" red } check_flag() { if is_set "$1"; then wrap_good "CONFIG_$1" 'enabled' else wrap_bad "CONFIG_$1" 'missing' fi } check_flags() { for flag in "$@"; do echo "- $(check_flag "$flag")" done } if [ ! -e "$CONFIG" ]; then wrap_warning "warning: $CONFIG does not exist, searching other paths for kernel config..." for tryConfig in "${possibleConfigs[@]}"; do if [ -e "$tryConfig" ]; then CONFIG="$tryConfig" break fi done if [ ! -e "$CONFIG" ]; then wrap_warning "error: cannot find kernel config" wrap_warning " try running this script again, specifying the kernel config:" wrap_warning " CONFIG=/path/to/kernel/.config $0 or $0 /path/to/kernel/.config" exit 1 fi fi wrap_color "info: reading kernel config from $CONFIG ..." white echo echo 'Generally Necessary:' echo -n '- ' cgroupSubsystemDir="$(awk '/[, ](cpu|cpuacct|cpuset|devices|freezer|memory)[, ]/ && $3 == "cgroup" { print $2 }' /proc/mounts | head -n1)" cgroupDir="$(dirname "$cgroupSubsystemDir")" if [ -d "$cgroupDir/cpu" -o -d "$cgroupDir/cpuacct" -o -d "$cgroupDir/cpuset" -o -d "$cgroupDir/devices" -o -d "$cgroupDir/freezer" -o -d "$cgroupDir/memory" ]; then echo "$(wrap_good 'cgroup hierarchy' 'properly mounted') [$cgroupDir]" else if [ "$cgroupSubsystemDir" ]; then echo "$(wrap_bad 'cgroup hierarchy' 'single mountpoint!') [$cgroupSubsystemDir]" else echo "$(wrap_bad 'cgroup hierarchy' 'nonexistent??')" fi echo " $(wrap_color '(see https://github.com/tianon/cgroupfs-mount)' yellow)" fi if [ "$(cat /sys/module/apparmor/parameters/enabled 2>/dev/null)" = 'Y' ]; then echo -n '- ' if command -v apparmor_parser &> /dev/null; then echo "$(wrap_good 'apparmor' 'enabled and tools installed')" else echo "$(wrap_bad 'apparmor' 'enabled, but apparmor_parser missing')" echo -n ' ' if command -v apt-get &> /dev/null; then echo "$(wrap_color '(use "apt-get install apparmor" to fix this)')" elif command -v yum &> /dev/null; then echo "$(wrap_color '(your best bet is "yum install apparmor-parser")')" else echo "$(wrap_color '(look for an "apparmor" package for your distribution)')" fi fi fi flags=( NAMESPACES {NET,PID,IPC,UTS}_NS DEVPTS_MULTIPLE_INSTANCES CGROUPS CGROUP_CPUACCT CGROUP_DEVICE CGROUP_FREEZER CGROUP_SCHED CPUSETS MACVLAN VETH BRIDGE NF_NAT_IPV4 IP_NF_FILTER IP_NF_TARGET_MASQUERADE NETFILTER_XT_MATCH_{ADDRTYPE,CONNTRACK} NF_NAT NF_NAT_NEEDED # required for bind-mounting /dev/mqueue into containers POSIX_MQUEUE ) check_flags "${flags[@]}" echo echo 'Optional Features:' { check_flags MEMCG_SWAP check_flags MEMCG_SWAP_ENABLED if is_set MEMCG_SWAP && ! is_set MEMCG_SWAP_ENABLED; then echo " $(wrap_color '(note that cgroup swap accounting is not enabled in your kernel config, you can enable it by setting boot option "swapaccount=1")' bold black)" fi } flags=( RESOURCE_COUNTERS CGROUP_PERF CFS_BANDWIDTH ) check_flags "${flags[@]}" echo '- Storage Drivers:' { echo '- "'$(wrap_color 'aufs' blue)'":' check_flags AUFS_FS | sed 's/^/ /' if ! is_set AUFS_FS && grep -q aufs /proc/filesystems; then echo " $(wrap_color '(note that some kernels include AUFS patches but not the AUFS_FS flag)' bold black)" fi check_flags EXT4_FS_POSIX_ACL EXT4_FS_SECURITY | sed 's/^/ /' echo '- "'$(wrap_color 'btrfs' blue)'":' check_flags BTRFS_FS | sed 's/^/ /' echo '- "'$(wrap_color 'devicemapper' blue)'":' check_flags BLK_DEV_DM DM_THIN_PROVISIONING EXT4_FS EXT4_FS_POSIX_ACL EXT4_FS_SECURITY | sed 's/^/ /' echo '- "'$(wrap_color 'overlay' blue)'":' check_flags OVERLAY_FS EXT4_FS_SECURITY EXT4_FS_POSIX_ACL | sed 's/^/ /' } | sed 's/^/ /' echo #echo 'Potential Future Features:' #check_flags USER_NS #echo