seanl's questions

I am having difficulty getting something that should be simple going. I have a Netgear DM111P modem connection to a broadband connection using PPPoA. Connected to this I have a Netgear FCG318 wireless router/8 port switch. This serves about a dozen users in a small office mostly wireless clients and some wired ones.

The modem is defaulted to act as a DHCP server when using PPPoA and if I have the DHCP disabled on the router nobody can connect to the internet as there not being assigned an IP. So the only way to connect is to specify an IP and everything works smoothly.

So the crux of the problem, I need to have DHCP enabled on the router so that users can connect but this seems to be causing the internet to fail within an hour of both modem and router being rebooted. What I suspect is that both are fighting for the right to assign IP's and eventually locking each other up as when the internet goes down I cant connect to either modem or router admin interfaces. Adding to this there seems to be an unusually large number of packet collisions on the network something like 3% of total traffic.

another curio seems to be the fact that they will not work when given the same class of IP ie currently the modem is 192.168.0.1 and the router 192.168.1.1 if i make them both 192.168.1.* respectively they refuse to cooperate.

So finally two questions, is the DHCP causing the router and modem to fight to the death and kill each other along with the internet connection and how to i resolve this.

What would be the correct IP address setting for the router and the modem given that I need the router to act as DHCP butcant disable the modem as the connection has to be PPPoA(its on the BT network in the UK).

I have Gitosis installed on a Centos box which I use to manage about a dozen repos(html/php/css etc) which I then manually pull through to live sites running on nginx/php as needed.

The issue I have is every time I do some work on one repository and then push my changes when I try and pull from another repo I get

ERROR:gitosis.serve.main:repository read access denied
Fatel: the remote end hung up unexpectedly

I can however over come this by editing the gitosis.conf file and removing the offending repo then push, re-edit and reinstate the repo and push, then going back to the repo and I can work away pushing and pulling fine.

My question is this why would working on one repo disrupt another and why would changing the gitosis.conf as discribed above fix this only temporarily and not permanently.

I can manage it the way it is fine its just inconvenient to go round the houses every time I need to switch repo etc

I have the following problem, I have a Centos 5.4 VPS hosting a few nginx sites (some static, some cakephp), I would like to be able to send email from each sites contact page through postfix to my google apps hosted email (different accounts for each site) so that apps can then send out an auto email to the person filling in the contact form etc

I have a bare-bones postfix installation with the following added into the main.cf config file. from using this guide

virtual_alias_domains = hash:/etc/postfix/virtual_alias_domains
virtual_alias_maps = hash:/etc/postfix/virtual_alias_maps

(both of these files have been converted into db files using postmap)

I have configured DNS correctly for each site and setup SPF records. (I'm aware R-DNS will still reference my actual hostname not the domain name and cause a possible spam issue but one thing at a time)

I can telnet localhost and the helo localhost so that I can send a command line email from an address in the virtual_alias_domains to an email in the virtual_alias_maps file which seems sends without giving an error but it is sending to my local linux account not the email address specified.

my question is am i approching this the wrong way in terms of the virtual alias mapping or is this even possible to do in the manner im trying. Any help is greatly appreciated thanks.

my postconf -n outlook looks like this

alias_database = hash:/etc/aliases
alias_maps = hash:/etc/aliases
command_directory = /usr/sbin
config_directory = /etc/postfix
daemon_directory = /usr/libexec/postfix
debug_peer_level = 2
html_directory = no
inet_interfaces = localhost
mail_owner = postfix
mailq_path = /usr/bin/mailq.postfix
manpage_directory = /usr/share/man
mydestination = $myhostname, localhost.$mydomain, localhost
myhostname = myactual hostname
mynetworks = 127.0.0.0/8
myorigin = $mydomain
newaliases_path = /usr/bin/newaliases.postfix
queue_directory = /var/spool/postfix
readme_directory = /usr/share/doc/postfix-2.3.3/README_FILES
sample_directory = /usr/share/doc/postfix-2.3.3/samples
sendmail_path = /usr/sbin/sendmail.postfix
setgid_group = postdrop
unknown_local_recipient_reject_code = 550
virtual_alias_domains = hash:/etc/postfix/virtual_alias_domains
virtual_alias_maps = hash:/etc/postfix/virtual_alias_maps

Hi somebody please help me out, I've asked this at stackoverflow as well but not got much of a response and was debating whether it was programming or server related.

I’m trying to setup a cakephp environment on a Centos server running Nginx with Fact CGI. I already have a wordpress site running on the server and a phpmyadmin site so I have PHP configured correctly.

My problem is that I cannot get the rewrite rules setup correct in my vhost so that cake renders pages correctly i.e. with styling and so on. I’ve googled as much as possible and the main consensus from the sites like the one listed below is that I need to have the following rewrite rule in place

location / {
  root /var/www/sites/somedomain.com/current;
  index index.php index.html;

  # If the file exists as a static file serve it 
  # directly without running all
  # the other rewrite tests on it
  if (-f $request_filename) { 
    break; 
  }
  if (!-f $request_filename) {
    rewrite ^/(.+)$ /index.php?url=$1 last;
    break;
  }
}

http://blog.getintheloop.eu/2008/4/17/nginx-engine-x-rewrite-rules-for-cakephp

problem is these rewrite assume you run cake directly out of the webroot which is not what I want to do. I have a standard setup for each site i.e. one folder per site containing the following folders log, backup, private and public. Public being where nginx is looking for its files to serve but I have cake installed in private with a symlink in public linking back to /private/cake/

this is my vhost

server {
    listen 80;
    server_name app.domain.com;

    access_log /home/public_html/app.domain.com/log/access.log;
    error_log /home/public_html/app.domain.com/log/error.log;

    #configure Cake app to run in a sub-directory
    #Cake install is not in root, but elsewhere and configured
    #in APP/webroot/index.php**

    location /home/public_html/app.domain.com/private/cake {
        index index.php;

        if (!-e $request_filename) {
            rewrite ^/(.+)$ /home/public_html/app.domain.com/private/cake/$1 last;
            break;
        }

    }

    location /home/public_html/app.domain.com/private/cake/ {
        index index.php;

        if (!-e $request_filename) {
            rewrite ^/(.+)$ /home/public_html/app.domain.com/public/index.php?url=$1 last;
            break;
        }

    }

    # pass the PHP scripts to FastCGI server listening on 127.0.0.1:9000
    location ~ \.php$ {
        fastcgi_pass   127.0.0.1:9000;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME /home/public_html/app.domain.com/private/cake$fastcgi_script_name;
        include        /etc/nginx/fastcgi_params;
    }

}

Now like I said I can see the main index.php of cake and have connected it to my DB but this page is without styling so before I proceed any further I would like to configure it correctly. What am I doing wrong……….

Thanks seanl