enachb

// sticking it in as part of the hadoop job
      FileSystem fs = FileSystem.get(jobConf);
      fs.mkdirs(new Path("/discovery/UrlDb"));
      fs.copyFromLocalFile(new Path(args[3] + "/gsb.blacklist"), new Path("/discovery/UrlDb"));
      fs.copyFromLocalFile(new Path(args[3] + "/gsb.malwarelist"), new Path("/discovery/UrlDb"));
      
      DistributedCache.addCacheFile(new URI("/discovery/UrlDb/gsb.blacklist"), jobConf);
      DistributedCache.addCacheFile(new URI("/discovery/UrlDb/gsb.malwarelist"), jobConf);

enachb

// Tell the RecordArchiver about our entries, so we can archive them later.
      RecordArchive ra = new RecordArchive(fs, inUrlsHDFS + "/../archive/incomingUrls");
      ra.addJobPath(new Path(inUrlsHDFS));
      Tap[] sourceFiles = new Tap[ra.getJobFiles().length];
      int i = 0;
      // Add only those files to the job, so we can accept more, while the job is running
      for (FileStatus f : ra.getJobFiles()) {
         LOG.info("Adding input files: " + f.getPath().getName());
         sourceFiles[i++] = new Hfs(new FlatfileSequenceFile(new Fields("key", "value")), f.getPath().toString());
         // sourceFiles[i++] = new Hfs(new TextLine(new Fields("value")), f.getPath().toString());

enachb

package com.bebo.hadoop.cascading.phishing;

import java.io.IOException;
import java.net.URISyntaxException;
import java.text.SimpleDateFormat;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Date;
import java.util.List;
import java.util.Vector;

enachb

import com.sun.istack.internal.Builder;

def writer = new StringWriter()  
writer.append '<?xml version="1.0" encoding="UTF-8"?>'
def builder = new groovy.xml.MarkupBuilder(writer) 

builder.urlset(xmlns:"http://www.sitemaps.org/schemas/sitemap/0.9") {
	url(){
		loc("http://sguys.com/MyManualLink1")
		changefreq("weekkly")

enachb

public class SiteMap {
		static void main(String[] args) { 
			
			def binding = new Binding();
			def products = [new Product(name:"ProdName1",productId:"P1000"), new Product(name:"ProdName2",productId:"P1001")]
			
			binding.setVariable "products", products			
			def shell = new GroovyShell(binding)			
		    String siteMap = shell.evaluate(new File('grails-app/conf/SiteMapConfig.groovy'))
			

enachb

#export GRAILS_VERSION="$(ls -lhr $HOME/.grails | egrep -i '1\.' | head -1 | gawk '{print $9 }')"
export GRAILS_VERSION=`cat $GRAILS_HOME/build.properties | grep "^grails.version=" | awk -F= '{ print $2 }' | tr -d '\r' | tr -d '\n'`

_get_domain_classes(){
	find ./grails-app/domain -iname *.groovy 2> /dev/null | tr \\n ' ' | sed 's/\.groovy//g' | sed 's/\.\/grails-app\/domain\///g' | tr '/' \.
}

_get_tests(){
	find ./test -iname *.groovy 2> /dev/null | sed 's/\.\/test\/integration\///g' | sed 's/\Tests.groovy//g' | tr '/' \.
}

enachb

import sys,json,fileinput

for line in fileinput.input():
  obj = json.loads(line)
  if obj["pageId"] == "AmericanIdol":
     print obj["utc"]

enachb

import sys,json,fileinput

for line in fileinput.input():
  obj = json.loads(line)
#  if obj["pageId"] == "AmericanIdol":                                       
  print "%s %s" % (obj["utc"],obj["pageId"])

enachb

import sys,json,fileinput
 
for line in fileinput.input():
  obj = json.loads(line)
#  if obj["pageId"] == "AmericanIdol":                                       
  print "%s %s" % (obj["pageUrl"], obj["time"])

----------------------------------------------------------------

enachb

---
- hosts: proxmox
  user: root
  tasks:
  - name: Add backports repo
    action: lineinfile dest='/etc/apt/sources.list.d/squeeze-backports.list' regexp='deb http://backports.debian.org/debian-backports squeeze-backports main' line='deb http://backports.debian.org/debian-backports squeeze-backports main' state=present create=true
  - name: Install collectd 5.1
    action: apt pkg=collectd-utils=5.1.0-3~bpo60+2 update-cache=yes state=installed default-release=squeeze-backports
  - name: Copy collectd.conf over
    action: copy src=qf/files/collectd.conf dest=/etc/collectd/collectd.conf owner=root group=root mode=644 backup=yes