#> cat /dev/diegor

Updates from January, 2011 Toggle Comment Threads | Keyboard Shortcuts

• 

  diegor 8:13 pm on September 19, 2011 Permalink | Reply
  Tags: howto ( 105 )   

  HOWTO: import old logs in AWStats 

  AWStats is a great software to analyse Apache logs for having statistics about your sites. Unfortunately it doesn’t offer a tool to import old logs into AWStats. For that I’ve created a small python script to do this. Here you go:

  import commands
  raw_data = commands.getoutput('ls -rt /var/log/apache2/wespot-access80.log.*.gz')
  logs = raw_data.splitlines()
  for log in logs:
      commands.getoutput('zcat %s > /tmp/tmplog' % log)
      commands.getoutput('/usr/lib/cgi-bin/awstats.pl -update -config=config.site.com -LogFile=/tmp/tmplog')

  If you have problems, comment this post!

  diegor and Andrea Corbellini are discussing. Toggle Comments

   
  • 

    Andrea Corbellini 11:09 pm on September 19, 2011 Permalink | Reply

    Hi! AWStats ships with a tool (logresolvemerge.pl) which does exactly the same thing. Also, I think the best way to use the script is not via a temporary file, but specifying something like this in the configuration:

    LogFile=”…/awstats-7.0/tools/logresolvemerge.pl …/access.log* |”

    The pipe symbol tells AWStats to read logs from the output of the command instead of from a file.

    [This is with AWStats 7.0, I don't know about older versions]

    • 

      diegor 10:51 am on September 20, 2011 Permalink | Reply

      Thanks Andrea for your suggestion! When I used AWStats, there wasn’t any tool to do that, so I created my own

  Reply Cancel reply

  Required fields are marked *

  Name *

  Email *

  Website

  

  Notify me of follow-up comments by email.

  Notify me of new posts by email.

• 

  diegor 6:06 pm on July 29, 2011 Permalink | Reply
  Tags: howto ( 105 )   

  HOWTO: rename all files in lowercase 

  Pretty simple.

  1. If you don’t have a *NIX operating system, please skip this post.
  2. Create a script with this content:

     #!/bin/sh
     for f in *; do
     g=`expr "xxx$f" : 'xxx\(.*\)' | tr '[A-Z]' '[a-z]'`
     mv -n "$f" "$g"
     done

     and call it “rename.sh”

  3. Give it execution permissions, typing:

     chmod u+x rename.sh

  4. Execute the script from the directory where you have your files to rename

  If you have two files, “Foo” and “foo“, you’ll be notified and no files will be overwritten. So, don’t worry about that, you won’t lose any information. To be sure, try it with a few files in “/tmp” directory.

  Question? Suggestion? Comment!

   

  diegor and Luca De Vitis are discussing. Toggle Comments

   
  • 

    Luca De Vitis 6:26 pm on July 29, 2011 Permalink | Reply

    ls -1 | while read file ; do mv -nv “${file}” “$(echo “${file}” | tr ‘[A-Z]‘ ‘[a-z]‘)” ; done

    • 

      diegor 6:53 pm on July 30, 2011 Permalink | Reply

      Hi Luca, thanks for the trick.
      I tried but it doen’t work.. I have this:

      “namefile not overwritten”

      It works only if I drop “-n” option from mv command (too risky!!)

• 

  diegor 3:14 pm on June 18, 2011 Permalink | Reply
  Tags: howto ( 105 )   

  HOWTO: postgresql on Dremhost Private Server 

  If you have a PS with DreamHost and you need to use PostgreSQL as database, you’ll see that’s impossible to login with postgres user and so you won’t able to use PostgreSQL.
  Dreamhost supports officially only MySQL, so they don’t provide any info about PostgreSQL.
  Anyway the problem of using PostgreSQL in DH’s PS is permit the login to postgres user. Follow this two steps:

  1. Become root, typing:

     #>sudo su

  2. Edit /etc/passwd file and find the line containing postgres. Change it from:

     postgres:x:32:32:PostgreSQL administrator,,,:/var/lib/postgresql:/bin/false

     to:

     postgres:x:32:32:PostgreSQL administrator,,,:/var/lib/postgresql:/bin/bash

  Now you can become postgres user and you can use PostgreSQL.
  Doubts? Comment!

  PS: Are you looking for a serious, riable and fast hosting? So Dreamhost is for you! Use Dreamhost promo code “DIEGOR” or click on this link to get 50$ off right now!

  jack1852 and diegor are discussing. Toggle Comments

   
  • 

    jack1852 5:16 pm on June 18, 2011 Permalink | Reply

    Domanda forse stupida: cos’è che viene cambiato rispetto a prima? (sappi che non ho un hosting su Dreamhost, solo curiosità linuxiana )

    • 

      diegor 11:17 pm on June 19, 2011 Permalink | Reply

      La cosa che cambia è il comando da eseguire al login. Ho editato il post. Grazie per la precisazione.

      • 

        jack1852 6:28 pm on June 20, 2011 Permalink | Reply

        In pratica abiliti la possibilità di login dell’utente postgres. Dovrebbe funzionare anche il comando (sempre da root): usermod -s /bin/bash postgres

• 

  diegor 9:45 am on May 20, 2011 Permalink | Reply
  Tags: howto ( 105 )   

  HOWTO: “ip_conntrack: table full, dropping packet” 

  I used to manage a server based on linux and once I had this strange message:

  #> ip_conntrack: table full, dropping packet

  One of the effects of this message is that you aren’t able to make and receive new connections.
  To solve that is enough increase ip_conntrack_max value. As first step check the current value typing:

  #> cat /proc/sys/net/ipv4/ip_conntrack_max
  65536

  Now, increase this value typing:

  #> echo 131072 > /proc/sys/net/ipv4/ip_conntrack_max

  Generally the right value of ip_conntrack_max is set to the total MB of RAM multiplied by 16. If you have 8GB RAM (8192 MB), the value should be 131072.
   
• 

  diegor 7:58 pm on March 4, 2011 Permalink | Reply
  Tags: howto ( 105 )   

  HOWTO: run cron job in seconds 

  If you know what cron is, you also know that the minimum time to execute a job is a minute. There is a small tip to execute a cron job every X seconds. For example if you need to run a cron job every 30 seconds (it’s a common case), you can do it adding these lines to crontab:

  * * * * * root sh /path/to/myscript.sh
  * * * * * root sleep 30 && sh /path/to/myscript.sh

  Both lines execute the job every minute, but the second one waits 30 seconds using “sleep” command from bash.

  If you have any question, comment this post.

  diegor and Andrea are discussing. Toggle Comments

   
  • 

    Andrea 4:01 pm on July 24, 2011 Permalink | Reply

    Very nice guide! It’s very tricky. Anyway do you know some implementation of a scheduler such cron that deals with seconds instead of minutes?

  • 

    diegor 11:17 pm on July 24, 2011 Permalink | Reply

    Mmm, sorry I don’t know any scheduler except cron..

• 

  diegor 7:53 pm on January 17, 2011 Permalink | Reply
  Tags: howto ( 105 )   

  HOWTO: fix offending key in SSH 

  When you perform an ssh connection to a remote machine, the 1st time you will be prompted to say ‘yes’ to authenticate the remote host.
  This feature is controlled by “StrictHostKeyChecking” parameter and it is set to yes by default. From security point of view, this option should be enable because protect you and your system against trojan horse attacks.
  Sometimes you need disabling it temporarily. To disable it you can do via console or via config file.

  Via console:

  #> ssh -o 'StrictHostKeyChecking no' username@remotemachine

  Via config file, adding below line to /etc/ssh_config (OSX) or /etc/ssh/ssh_config (linux):

  StrictHostKeyChecking no

  Once you turned it on, you can solve “offending key” deleting it from “~/.ssh/know_hosts” through this terminal command:

  # sed -i '19d' ~/.ssh/known_hosts

  where 19 is the line containing offending key.

  If you have suggestions, troubles or you feel alone, comment this post!