Ask Cerebro

DEPRECATED: askcerebro will shut down on 04/26/14.

What ?

Cerebro is an API-like endpoint that returns information about a package.

Releases type, versions and urls are returned as a JSON so it can be used for rolling upgrades of common software packages.

It's aimed for people who like to compile from sources as it always returns the source.tar.gz of packages.

You feed it disks and it tells you everything about them.

Why ?

I got tired having to check for the latest version of Nginx, Redis, etc and switching variables in Chef, Fabric, Capistrano and wherever you can think.

For example, looking at the Logstash cookbook from Lusis and contributors:

node['logstash']['agent']['version'] - The version of Logstash to install. Only applies to jar install method.

Logstash agent ['version'] can now be easily updated when a new release is available.

I needed something easy, fast, easy to maintain and was using it personally with restricted IPs. A friend asked me if he could used it as well, so I just removed all IPs restrictions and put it on Github.

Where ?

Just make a call to askcerebro.com/package where package is the package you want to get information about.

How ?

JSON API

Here is an example:

$> curl -i www.askcerebro.com/redis

Answer should be 200 OK if found with a JSON answer:

HTTP/1.1 200 OK
Date: Tue, 12 Mar 2013 18:50:59 GMT
Content-Type: text/plain; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive

{
  "redis": {
    "releases": [
      "stable",
      "legacy"
    ],
    "stable": {
      "md5": "c4be422013905c64af18b1ef140de21f",
      "sha256": "3b9439636c58ca06bee538a0f7298e02a33fcf98b8fa845c0b0cf8567751e948",
      "url": "http://redis.googlecode.com/files/redis-2.6.13.tar.gz",
      "version": "2.6.13"
    },
    "legacy": {
      "md5": "13b9955a924be15b5fe67f970a5386b5",
      "sha256": "d71b6372f42fcbdc77a9601f1dd6a029ed57f7f77ac3b18bfed8670fb8c74697",
      "url": "http://redis.googlecode.com/files/redis-2.4.18.tar.gz",
      "version": "2.4.18"
    }
  }
}

If a package is not available in Cerebro (for example perl), you'll get a 404 HTTP NOT FOUND with an "NA" answer:

$> curl -i www.askcerebro.com/perl
	
HTTP/1.1 404 Not Found
Date: Tue, 12 Mar 2013 18:51:35 GMT
Content-Type: text/plain; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive

NA

If you're trying to hit an endpoint unknown to Cerebro, you'll get a 404 HTTP NOT FOUND answer:

$> curl -i www.askcerebro.com/shady/endpoint 

HTTP/1.1 404 Not Found
Date: Tue, 12 Mar 2013 18:51:35 GMT
Content-Type: text/plain; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive

If Cerebro is down, you'll get a 503 Service Temporarily Unavailable:

$> curl -i www.askcerebro.com/down

HTTP/1.1 503 Service Temporarily Unavailable
Date: Wed, 04 Dec 2013 23:41:28 GMT
Content-Type: text/plain; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive

HUMAN API

You can directly get the plain text answer about a package, depending of the release type.

Currently only version, url, md5 and sha256 are supported.

Here is an example:

$> curl -i www.askcerebro.com/mongodb/nightly/url

HTTP/1.1 200 OK
Date: Tue, 12 Mar 2013 19:02:25 GMT
Content-Type: text/plain; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive

http://github.com/mongodb/mongo/tarball/v2.2-latest

And another:

curl -i www.askcerebro.com/redis/legacy/version

HTTP/1.1 200 OK
Date: Tue, 12 Mar 2013 19:03:35 GMT
Content-Type: text/plain; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive

2.4.18

How to add a package to Cerebro ?

Just add another disk with the following YAML structure:

new_package:
	releases: ['stable', 'development', 'legacy']
	stable:
		version: 1.2.7
		url: http://example.com/download/stable-1.2.7.tar.gz
		md5: abcdefghijkl124345
		sha256: d71b6372f42fcbdc77a9601f1dd6a029ed57f7f77ac3b18bfed8670fb8c74697
	development:
		version: 1.3.14
		url: http://github.example.com/download/latest.tar.gz
		md5: abcdefghijkl124345
		sha256: d71b6372f42fcbdc77a9601f1dd6a029ed57f7f77ac3b18bfed8670fb8c74697
	legacy:
		version: 1.0.15
		url: http://example.com/download/legacy-1.0.15.tar.gz
		md5: abcdefghijkl124345
		sha256: d71b6372f42fcbdc77a9601f1dd6a029ed57f7f77ac3b18bfed8670fb8c74697

Once I merge your disk, you should be able to ask Cerebro about it within couple mins.

How to use it with Chef ?

You could technically use it for rolling upgrades (be careful), with something like this (YMMV):

require "net/http"
require "uri"
require "json"
require 'chef/log'

begin
  redis_info = URI.parse("http://www.askcerebro.com/redis")
  info = JSON.parse(Net::HTTP.get_response(redis_info).body)
  redis_version = info['redis']['stable']['version']
  node.default[:redis][:server][:version] = redis_version.chomp
rescue => e
  Chef::Log.info("Cerebro is down, error is #{e}")
  Chef::Log.info("Falling back to default attribute version.")
  node.default[:redis][:server][:version] = "2.6.10"
end

node.default[:redis][:server][:addr]    = "127.0.0.1"

Please also look at the ruby example.

Need Help or Want to Contribute ?

All contributions are welcome: ideas, patches, documentation, bug reports, complaints, and even something you drew up on a napkin.

It is more important to me that you are able to contribute and get help if you need it.

That said, some basic guidelines, which you are free to ignore :)

• Have a problem you want askcerebro to solve for you ? You can email me personally ([email protected])
• Have an idea or a feature request? File a ticket on Github, or email me personally ([email protected]) if that is more comfortable.
• If you think you found a bug, it probably is a bug. Please file a ticket on Github.
• If you want to add new disks, best way is to fork this repo and send me a pull request. If you don't know git, I also accept diff(1) formatted patches - whatever is most comfortable for you.

Programming is not a required skill. Whatever you've seen about open source with maintainers or community members saying "send patches or die" - you will not see that here.

Contributors

Thank you for contributing !