Get the regions by name for ec2 connection Also a script to rsync with other nodes in starcluster using the alias tag.

get_region.py

import boto.ec2
def get_region(region_name):
    for region in boto.ec2.regions():
        if region_name == region.name:
            return region
            
conn = boto.connect_ec2(region=get_region( 'us-east-1'))

rsync.py

#!/usr/bin/env python                                                           
import sys                                                                      
import boto                                                                     
import boto.ec2                                                                 
import boto.utils                                                               
import re                                                                       
import logging                                                                  
import subprocess                                                               
def get_region(region_name):                                                    
    for region in boto.ec2.regions():                                           
        if region_name == region.name:                                          
            logging.info("node region[%s]" % region.name)                       
            return region                                                       
                                                                                
def get_my_alias():                                                             
    identity = boto.utils.get_instance_identity()                               
    conn = boto.connect_ec2(region=get_region( identity['document']['region'])) 
    mytags = conn.get_all_tags(filters={'resource-id':identity['document']['instanceId'], 'key':'alias'})
    for tag in mytags:                                                          
        if tag.name == 'alias':                                                 
            logging.info("my alias[%s]" % tag.value)                            
            return tag.value                                                    
                                                                                
def get_nodes():                                                                
    nodes = []                                                                  
    my_alias = get_my_alias()                                                   
    identity = boto.utils.get_instance_identity()                               
    conn = boto.connect_ec2(region=get_region( identity['document']['region'])) 
    m = re.match(r'(?P<cluster>\w*-?)(master|node\d\d\d)', my_alias)            
    if m:                                                                       
        mycluster = m.groupdict()['cluster']                                    
        aliases = conn.get_all_tags(filters={'resource-type':'instance', 'key':'alias'})
        for alias in aliases:                                                   
            my_str = mycluster + r'((master)|(node\d\d\d))'                     
                                                                                
            if re.match(my_str, alias.value) and not alias.value == my_alias :  
                nodes.append(alias.value)                                       
    logging.info( "Cluster nodes [%s]" % ' ,'.join( nodes ))                    
    return nodes                                                                
                                                                                
def rsync_home( nodes ):                                                        
    for node in nodes:                                                          
        node_str = 'sgeadmin@%s:~/' % node                                      
        command = ['rsync', '-avz', '~/', node_str]                             
        logging.info("Command: %s" % ' '.join(command))                         
        result = subprocess.check_output(' '.join(command), shell=True)         
        logging.info("Response:")                                               
        logging.info("=" * 50)                                                  
        logging.info(result)                                                    
if __name__ == '__main__':                                                      
    logging.basicConfig(filename='rsync.log',level=logging.DEBUG,               
            format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')