** 写了一个处理音视频的服务， 之前都是直接安装调用，每次新增服务器都要重新装一遍比较麻烦，所有有了这篇文章， 本文将基于HYPERF的基础DOCKERFILE 进行测试及搭建: HYPERF、FFMPEG、SPEEX2WAV环境 **
构建一个未瘦身的环境 dockerfile如下
FROM hyperf/hyperf:latest
LABEL maintainer="LW放下" version="1.0" license="MIT"
COPY ffmpeg-4.1 /tmp/ffmpeg-4.1
COPY lame-3.99.5 /tmp/lame-3.99.5
COPY speex-1.2.0 /tmp/speex-1.2.0
COPY wechat-speex-declib /tmp/wechat-speex-declib
RUN set -ex
&& ls -al
&& apk add --no-cache --update autoconf
automake
bash
binutils
bzip2
cmake
curl
coreutils
diffutils
file
g++
gcc
gperf
libtool
make
tar
yasm
nasm
zlib-dev
expat-dev
&& cd /tmp/lame-3.99.5
&& ./configure
&& make
&& make install
&& cd /tmp/ffmpeg-4.1
&& ./configure --enable-libmp3lame
&& make
&& make install
&& cd /tmp/speex-1.2.0
&& ./configure
&& make
&& make install
&& cd /tmp/wechat-speex-declib
&& make
&& cp ./bin/speex_decode /usr/local/bin/speex2wav
&& apk del --purge *-dev
&& rm -rf /var/cache/apk/* /tmp/*
&& php -v
&& php -m
&& echo -e "33[42;37m Build Completed :).33[0m
"
WORKDIR /hyperf-skeleton/
所需文件可以查看 speex2wav ffmpeg
如下为build出来的image,可以发现很大(大得惊人)
开始骚操作找到相关依赖及执行文件位置，需要掌握如下技能
- 百度
- linux基础命令： cp、ls、ln、which、ldd、grep
- docker基础
下面命令不知道干啥的自己百度
ffmpeg相关文件
基于刚刚的image创建一个容器
docker run --name hyperf-fmpeg-test -p 9420:9420 -it --entrypoint /bin/sh hyperf-fmpeg-test:1.0
which ffmpeg
/usr/local/bin/ffmpeg
ldd ffmpeg
/lib/ld-musl-x86_64.so.1 (0x7efd831c9000)
libz.so.1 => /lib/libz.so.1 (0x7efd81741000)
libmp3lame.so.0 => /usr/local/lib/libmp3lame.so.0 (0x7efd81690000)
libc.musl-x86_64.so.1 => /lib/ld-musl-x86_64.so.1 (0x7efd831c9000)
ldd /usr/local/bin/speex2wav
/lib/ld-musl-x86_64.so.1 (0x7f86c0e28000)
libspeex.so.1 => /usr/local/lib/libspeex.so.1 (0x7f86c0e08000)
libc.musl-x86_64.so.1 => /lib/ld-musl-x86_64.so.1 (0x7f86c0e28000)
查看具体位置及文件
ls -al /lib/ | grep ld-musl
-rwxr-xr-x 1 root root 584304 Aug 7 2019 ld-musl-x86_64.so.1
lrwxrwxrwx 1 root root 19 Nov 6 01:35 libc.musl-x86_64.so.1 -> ld-musl-x86_64.so.1
ls -al /lib/ | grep libz
lrwxrwxrwx 1 root root 14 May 9 2019 libz.so.1 -> libz.so.1.2.11
-rwxr-xr-x 1 root root 100144 Nov 15 2018 libz.so.1.2.11
ls -al /usr/local/lib/ | grep libmp3
-rw-r--r-- 1 root root 749162 Apr 24 09:30 libmp3lame.a
-rwxr-xr-x 1 root root 943 Apr 24 09:30 libmp3lame.la
lrwxrwxrwx 1 root root 19 Apr 24 09:30 libmp3lame.so -> libmp3lame.so.0.0.0
lrwxrwxrwx 1 root root 19 Apr 24 09:30 libmp3lame.so.0 -> libmp3lame.so.0.0.0
-rwxr-xr-x 1 root root 559128 Apr 24 09:30 libmp3lame.so.0.0.0
如上所示，找到了文件我们得拿出来放到基础的hyperf环境里面
1.先拷贝所需文件到宿主机
docker cp 9aff3c919e19:/lib/ld-musl-x86_64.so.1 /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/lib/libz.so.1.2.11 /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/lib/libmp3lame.a /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/lib/libmp3lame.la /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/lib/libmp3lame.so.0.0.0 /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/lib/libspeex.a /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/lib/libspeex.la /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/lib/libspeex.so.1.5.1 /www/wgc_service/ffmpeg-static/
docker cp 9aff3c919e19:/usr/local/bin/speex2wav /www/wgc_service/ffmpeg-static/
开始测试
创建基础hyperf容器
docker run -v /www/wgc_service/ffmpeg-static/:/www/wgc_service/ffmpeg-static/ -p 9404:9404 --name hyperftestnew -it --entrypoint /bin/sh hyperf/hyperf:latest
进入容器执行(其实就是还原另一个容器的一些文件路径及结构)
cp -p /www/wgc_service/ffmpeg-static/ld-musl-x86_64.so.1 /lib
cp -p /www/wgc_service/ffmpeg-static/libz.so.1.2.11 /lib
cp -p /www/wgc_service/ffmpeg-static/libmp3lame.a /usr/local/lib
cp -p /www/wgc_service/ffmpeg-static/libmp3lame.la /usr/local/lib
cp -p /www/wgc_service/ffmpeg-static/libmp3lame.so.0.0.0 /usr/local/lib
cp -p /www/wgc_service/ffmpeg-static/libspeex.a /usr/local/lib
cp -p /www/wgc_service/ffmpeg-static/libspeex.la /usr/local/lib
cp -p /www/wgc_service/ffmpeg-static/libspeex.so.1.5.1 /usr/local/lib
cp -p /www/wgc_service/ffmpeg-static/ffmpeg /usr/local/bin
cp -p /www/wgc_service/ffmpeg-static/speex2wav /usr/local/bin
ln -s /usr/local/lib/libmp3lame.so.0.0.0 /usr/local/lib/libmp3lame.so.0
ln -s /usr/local/lib/libmp3lame.so.0.0.0 /usr/local/lib/libmp3lame.so
ln -s /usr/local/lib/libspeex.so /usr/local/lib/libspeex.so.1.5.1
ln -s /usr/local/lib/libspeex.so.1 /usr/local/lib/libspeex.so.1.5.1
来吧， 测试一下看看行不行(注意替换路径)
ffmpeg -version
speex2wav -version
ffmpeg -i input.mp3 -af volume=2 -y out.mp3
不出意外是可以的，开始编写dockerfile吧
FROM hyperf/hyperf:latest
LABEL maintainer="LW放下" version="1.0" license="MIT"
COPY ld-musl-x86_64.so.1 /lib
COPY libz.so.1.2.11 /lib
COPY libmp3lame.a /usr/local/lib
COPY libmp3lame.la /usr/local/lib
COPY libmp3lame.so.0.0.0 /usr/local/lib
COPY ffmpeg /usr/local/bin
COPY libspeex.a /usr/local/lib
COPY libspeex.la /usr/local/lib
COPY libspeex.so.1.5.1 /usr/local/lib
COPY speex2wav /usr/local/bin/
RUN set -ex
&& chmod +x /usr/local/bin/ffmpeg
&& chmod +x /usr/local/bin/speex2wav
&& ln -s /usr/local/lib/libmp3lame.so.0.0.0 /usr/local/lib/libmp3lame.so.0
&& ln -s /usr/local/lib/libmp3lame.so.0.0.0 /usr/local/lib/libmp3lame.so
&& ln -s /usr/local/lib/libspeex.so.1.5.1 /usr/local/lib/libspeex.so
&& ln -s /usr/local/lib/libspeex.so.1.5.1 /usr/local/lib/libspeex.so.1
&& ffmpeg -version
&& speex2wav -version
&& php -v
&& php -m
&& echo -e "33[42;37m Build Completed :).33[0m
"
WORKDIR /hyperf-skeleton/
构建试试
docker build -t hyperf-fmpeg-base:1.0 -f Dockerfile .
如图所示, 可以发现少了几百M
