{"id":401,"date":"2016-11-01T00:36:35","date_gmt":"2016-10-31T16:36:35","guid":{"rendered":"http:\/\/finaldie.com\/blog\/?p=401"},"modified":"2017-08-27T20:48:07","modified_gmt":"2017-08-28T03:48:07","slug":"skull-engine-getting-started","status":"publish","type":"post","link":"https:\/\/finaldie.com\/blog\/skull-engine-getting-started\/","title":{"rendered":"Skull-Engine: Getting Started"},"content":{"rendered":"

\"Halloween\"<\/a><\/p>\n

After introducing Skull<\/em><\/a>, this time let’s see how to install and use Skull<\/em><\/a> to create a Skull<\/em> project.<\/p>\n

Install Skull<\/em><\/h2>\n

As mentioned in Introduction<\/strong>, Skull<\/em> includes 3 majors parts: skull-core<\/em>, skull-apis<\/em> and skull-scripts<\/em>. To play with it, we need to install Skull<\/em> first, then use Skull<\/em> commands to create\/manage a Skull<\/em> project.<\/p>\n

And as we expected, Skull<\/em> can not only work in the big data center, but also in the resource very limited environment(embedded device), like Raspberry Pi<\/strong>, follow me to see how to do that :)<\/p>\n

<\/p>\n

Install Dependencies<\/h3>\n

\"datacenter\"<\/a><\/p>\n

In generally Skull<\/em> is working in the Data Center<\/em>, with powerful x86 CPUs and enough RAM. For example, here we use Ubuntu14.04<\/em> as an example.<\/p>\n

sudo apt-get install autoconf valgrind expect libyaml-dev python-dev python-pip libprotobuf-dev protobuf-compiler libprotobuf-c0-dev protobuf-c-compiler\n\nsudo pip install PyYAML protobuf pympler WebOb\n\ngit clone git@github.com:finaldie\/skull.git\ncd skull\ngit submodule update --init --recursive\nmake dep\nsudo make install-dep\n<\/code><\/pre>\n
Build and Install<\/h3>\n
make -j4\nsudo make install\n<\/code><\/pre>\n
After that, Skull<\/em> has installed into the system, type skull<\/strong><\/span> we would see:<\/p>\n
final@ubuntu1404: ~>skull\nSkull 0.9.10 - A fast to start, easy to maintain, high productive serving framework.\nMore information see https:\/\/github.com\/finaldie\/skull\n\nusage:\n\\_ skull create\n\\_ skull config\n\\_ skull common\n\\_ skull workflow\n\\_ skull module\n\\_ skull service\n\\_ skull build\n\\_ skull deploy\n\\_ skull start\n\\_ skull version\n\\_ skull help\n<\/code><\/pre>\n
Now, Skull<\/em> is ready to play with it then :)<\/p>\n
Install Skull<\/em> in Raspberry Pi<\/em><\/h2>\n
\"raspberry<\/a><\/p>\n
Above shows Skull<\/em> is working fine in normal Linux<\/em> releases, but in the embedded device, like Raspberry Pi<\/strong>, the situation is totally different. It only has a ARMv6<\/em> processor(32bit) with only 384MB RAM(Model B), the good news is, Skull<\/em> is totally 32bit compatible, it’s working perfectly in this environment, let’s see how to fit Skull<\/em> into Raspberry<\/em> :)<\/p>\n
To make it works, we have to disable the jemalloc<\/strong> memorypool, since it has some compiling problems in this environment, the rests are simple and easy:<\/p>\n
# Install dependencies packages\nsudo apt-get install autoconf valgrind expect libyaml-dev python-dev python-pip libprotobuf-dev protobuf-compiler libprotobuf-c0-dev protobuf-c-compiler\n\nsudo pip install PyYAML protobuf pympler WebOb\n\n# Clone Skull\ngit clone git@github.com:finaldie\/skull.git\ncd skull\ngit submodule update --init --recursive\n\n# Build Skull dependency libraries\nmake dep disable_jemalloc=true\nsudo make install-dep\n\n# Build Skull itself\nmake -j4 disable_jemalloc=true\nsudo make install\n<\/code><\/pre>\n
OK, right now we are all set, type skull<\/strong> to verify the Skull<\/em> commands are working fine here.<\/p>\n
Basic flow<\/h2>\n
After installing Skull<\/em>, let’s create a basic Skull<\/em> project, which only need few steps:<\/p>\n
    \n
  • skull create project<\/code><\/li>\n
  • skull workflow --add<\/code><\/li>\n
  • skull module --add<\/code><\/li>\n
  • skull service --add<\/code> (optional)<\/li>\n<\/ul>\n
    OK, here we created an simple example project with one default module with it:<\/p>\n