I decided to put together a utility script to allow me to connect to my Amazon
EC2 instance conveniently from the command line. While this started out as a few
lines at first (following the very helpful AWS manuals) I soon realized that I
may want to incorporate several commands (such as start/stop/describe) as
input options and then perform verification checks upon establishing a
connection.
To do these 'correctly', more or less, would require a module to store all my
AWS-relevant configuration settings. This, coupled with use of my own
environment's key system parameter-argument couplings and logging utility, gives
the appearance that the code is somewhat more bloated than it really is, but
it works, nevertheless.
**Note - Some Editing Required: I highlight optional/environment-specific
components in the script descriptions below.
1). awsConn.pl - main script for managing connections.
This script imports the standard Perl modules: POSIX, File, & Time
and 3 modules relevant to my own environment: FsSysVars, FsLog, and
CfgAws. The first two are responsible for setting certain
environment variables and logging tasks executed in a standardized
way, respectively. These maybe either replaced or ignored so long
as their associated instantiation tasks within the code body are
also edited to negate or modify their use.
The module CfgAws, however, is crucial to the operation of this
script as it keeps a record of the details pertinent to EC2
instances. These include the instance id, for example, which
is of course required when attempting to connect to an AWS server.
Sample usage is as follows:
awsConn.pl --argv cmd=start
awsConn.pl --argv cmd=connect
awsConn.pl --argv cmd=describe
awsConn.pl --argv cmd=stop
2). CfgAws.pm - module which contains configuration settings for one or more
AWS instances (in my case, EC2 instances). This module contains
important information like the path and filename of key
authentication and authorization files such as the AWS ssh key file
name and location associated with a given instance as well a set of
higher level variables such as AWS status codes and definitions
(gleaned from AWS manuals) and account information and certain other
operation-specific file settings.
Conveniently, this module is structured such that different
instances can be given different names in the local context (e.g.
'dev' versus 'prod') corresponding to different EC2 instances with
their own relative configuration settings (instance id number, etc.).
Unlike the other two modules, this one is required. Obviously, the
path and/or name/structure maybe modified to match your environment
so long as key machine instance identifiers are passed to the
awsConn.pl script.
3). sshaws.bash - wrapper for awsConn.pl.
Simplifies calls to awsConn.pl by allowing the user to simply type:
sshaws -i <instanceName> -c <command>
where:
instanceName: an instance name parameter specified in the
configurations module with associated connection and other descriptive
details.
command: choose from start/stop/describe/connect
A key way to effectuate such a call would be to include a function
in your .bashrc profile or one of your environment profile scripts
and define a function sshaws which calls sshaws.bash (or just use
'shaws.bash' if you don't mind typing that much).
Hope you find these useful!