bobinabottle's questions

I'm a long time postgres user, but am now looking after maintenance for mysql databases. One thing that I'm used to from postgres, is when you run a sql script it will output on the command line what it is executing.. i.e.

cat myscript.sql | psql mydatabase
BEGIN;
INSERT;
COMMIT;

When I do a similar thing with mysql, I get no output and it will just churn away without giving me any status

cat myscript.sql | mysql mydatabase

Are there any flags I can use or something to the mysql binary so I can get an idea of whats going on when the script is executing?

I have a somewhat unique situation. Our systems team manages our main puppet master, and the development team is fine with everything however they are thinking of using it to control some elements on their desktop machines, whilst still being connected to our central puppet master.

Since we don't want the changes they make to go into our puppet master.. is there a way of puppet importing a manifest from the node directly?

As in.. on the developer machine, they put a file "/root/development.pp" or something, and then on our puppet master we put something like

node { "developermachine":
  # Do the majority of normal things
  # import "/root/development.pp"
}

We have a few different options we can take about security of write access to the puppet manifests, but if puppet were to support something like this it would probably be the cleanest for us.

Any help is appreciated :)

In apache2, I can see all the active virtual hosts by doing apache2ctl -S. This is useful since I don't need to plough through the config files, and also if I have made changes to the config files and not yet restarted apache2.

Is there an equivalent function in nginx?

I have a definition in my puppet manifests called postgresql::user

I have put in the following, to make sure puppet doesn't try to do something twice:

if !defined (Postgresql::User["dbuser"]) { 
        postgresql::user { "dbuser": 
            ensure => present
        } 
    }

However I am still getting the errors in puppetd --test

err: Could not retrieve catalog: Duplicate definition: Postgresql::User[dbuser] is already defined in file /etc/puppet/modules/dbserver/manifests/postgresql8-3.pp at line 22; cannot redefine at /etc/puppet/modules/dbserver/manifests/postgresql8-3.pp:7 on node mynode

I have a feeling it doesn't like the capitalisation and `::' when checking if defined. Is there way to escape this?

I can understand according to http://docs.puppetlabs.com/guides/language_guide.html that I can pass an array to a definition.

define php::pear() {
        package { "`php-${name}": ensure => installed }
}

php::pear { ['ldap', 'mysql', 'ps', 'snmp', 'sqlite', 'tidy', 'xmlrpc']: }

However, is there a way that I can pass more than one array to do the same thing? For instance, if I wanted some pear libraries installed, and some absent, could I do this?

define php::pear($ensure) {
    package { "`php-${name}": ensure => $ensure }
}

php::pear { ['ldap', 'mysql', 'ps', 'snmp', 'sqlite', 'tidy', 'xmlrpc']: 
    ensure => ['installed', 'installed', 'absent', 'absent', 'installed', 'installed', 'installed']
}

Firstly, is something like this possible, or are there better ways of doing something like this with multidimensional arrays or something?

Any help would be appreciated, I'm finding the puppet documentation lacking a little in this area.

Cheers

Is there an equivalent of gitosis for subversion? As in a wrapper package that allows ssh access via ssh keys, with everyone using the one ssh user?

Question says it all really, how can I run postgres 7.4 on debian lenny? Our developers don't have the resource right now to modify a legacy application we have to work with pg 8.3, and debian etch is no longer supported.

We currently use a number of linux routers, and currently a number of the team tend to jump into servers and change the routing rules when needed, and the auditing is becoming incredibly tedious.

For all other server configs we store in a git repository, and are migrating to using puppet. I haven't been able to figure out how to manage the iproute2 configuration however, as all changes are done live.

This was a similar question for cisco routers: Managing and auditing existing Cisco firewall access control and routing rules I'm wondering if something like this can be done in linux iproute2?

I'm trying to get the simpletest module for drupal to work, which relies on php-curl to fetch its requests.

When I as the user root do the following it works fine:

root@server:~# php -a
Interactive mode enabled

<?php

$ch = curl_init("http://testsite.drupal.dev/user");
$fp = fopen("example.html", "w");

curl_setopt($ch, CURLOPT_FILE, $fp);
curl_setopt($ch, CURLOPT_HEADER, 0);

curl_exec($ch);
curl_close($ch);
fclose($fp);
?>

root@server:~# cat /root/example.html 
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<!-- Blah Blah -->
</body>
</html>
root@server:~# 

However when the www-data user does it, it fails:

root@server:~# su - www-data
www-data@server:~$ php5 -a
Interactive mode enabled

<?php

$ch = curl_init("http://testsite.drupal.dev/user");
$fp = fopen("example.html", "w");

curl_setopt($ch, CURLOPT_FILE, $fp);
curl_setopt($ch, CURLOPT_HEADER, 0);

curl_exec($ch);
curl_close($ch);
fclose($fp);
?>

www-data@server:~$ cat example.html 
www-data@server:~$ 

Has anyone seen this before? I'm struggling to see how I can fix this. Any help is appreciated! :)

Environment: Debian Etch / Apache2 / php5

I have 3 RAID arrays in my server, mapped to 2 PVs in my lvm2 setup.

apollo:~# pvs
  PV         VG      Fmt  Attr PSize PFree
  /dev/sda1  apollo  lvm2 a-   3.64T    0 
  /dev/sda2  apollo  lvm2 a-   1.14T    0 
  /dev/sdb   apollo1 lvm2 a-   2.05T    0 
apollo:~# lvs
  LV        VG      Attr   LSize   Origin Snap%  Move Log Copy% 
  lvol0     apollo  -wi-ao   4.77T                              
  ling-ling apollo1 -wi-ao 895.41G                              
  lvol0     apollo1 -wi-ao   1.17T                              
  lvol1     apollo1 -wi-a-  32.00M                              

I want to remove the LV "ling-ling" to free up some space, however I want to expand lvol0 with that free space, which is residing on the other VG.

Is there a way in LVM to allocate a percentage of a PV to different VGs without repartitioning? The only solution I can think of is to repartition /dev/sdb but I would prefer not to, since this is a live server.

Thanks :)

We have a number of external systems reading from our emails, and some are tripping over others. Basically what I need to do is replace an email subject like:

[RT #179497] Internal: blah blah

with

[RT 179497] Internal: blah blah

So drop the hash. Does anyone know the regex to achieve this with postfix header_checks?

Our exchange 2k3 database got corrupted and we couldn't revive it, so we're going to restore the last backup, which was unfortunately a week ago :-S

My question is, what happens to the cached emails on users machines for that week? We have outlook and Entourage clients connecting to it, with some imap users via other clients.

As soon as the server comes alive again, will the cached emails on the users machines be deleted? Or synched up to the server? Or will nothing happen?

Thanks :) Just wondering if anyone had experience with this.

The last guy forgot to enable vnc connections on our Xen blade and I need to install a windows domain, is there any way I can enable vnc without shutting down half our company?

From what I can tell I need to edit xend-config.sxp which would likely require a restart of xend, which means shutting down all the running domains. Is there any way around this?

thanks :)

Our web application is getting more and more traffic, which is making our poor pg8.3 database server have a little trouble keeping up. I've had a look into using pgpool II for clustering the db to relieve a little strain, and I was wondering how this should be done to minimise downtime considering I would be clustering a live database.

Has anyone had experience with this or know of any guides to follow? Cheers :)

I have just started at a new company and I am inspecting their current server config. The HP 480c blades in a c7000 chassis aren't responding to ILO, although the chassis ILO is working fine. I have a feeling the last sysadmin configured the blades ILO as static IPs and it is not responding correctly. The servers are sitting in a datacenter and I'm hoping to be able to fix this remotely.

Is there a way that I can change the ILO static IPs for the blades remotely? If not and I do have to go onsite, how do I change the IP addresses of the ILO for the blades? (Sorry I'm not very familiar with HP servers)

thanks for you help!

This is somewhat related to this question but is a different question.

We have a central Hg repository, served to users via SSH and mercurial-server. We have a number of Mac, Linux and Windows clients connecting to it.

It has happened twice now that one of the Windows users has corrupted their repository, then pushed back to the central one corrupting it. I want to write an incoming hook script on the central repository to prevent a transaction from being accepted if it will corrupt the central repo.

Although unfortunately I don't know enough about Mercurial to write such a script. Any possibility that someone else has come across this? Personally I'm not quite sure why hg doesn't do this by default.

I want to be able to print the number of lines from bash as such: Line number (as counted from the top) -> end of the file

It seems tail will only count lines from the bottom.

Does anyone know how to do this? Thanks.

I've tried the following

# Where $1 is the file I'm reading in

# Get line number of error:
LINENUM=$( grep -n "$LAST_ERROR_DATE" $1 | egrep $LOG_THRESHOLD | grep $LAST_HOUR: | sed 's/:/ /g' | awk '{print $1}' | head -n 1 )

echo $LINENUM

# This returns 995

# Print everything from linenumber downwards
MESSAGE=$( awk 'NR >= $LINENUM' $1 )

This works when I manually pop in 995 to awk instead of $LINENUM, it doesn't seem to read in my bash variable. Any ideas?

I have a unique situation I think, google isn't telling me much.

I want to have a way to have a hyperlink in html to simply go to: http://:8081 http://:8080

When the page itself is running on port 80.

This would usually be easy, however this is being developed as a virtual appliance to run offline on our sales teams machines, and as such the IP address will always be different dependant on the machine its running on. I also cannot use DNS for this.

Afaik, it is impossible to have a relative hyperlink to link to another port. I then thought I could perhaps just link to "./site1" and "./site2" and then use mod_rewrite to write those links to the right port.

Does anyone know how to do this? I went do it and got stuck when I realised I wanted to rewrite the URL to a static IP which I can't do! (apache or nginx rewrite rules, or any other solution very welcome!)

Has anyone heard any announcements from vmware as to when these will be officially supported? Their website is a bit of a mess. I'm wondering if ESXi 3.5 will support them or ESXi4 ?

Googling around only provides info about installing the RC version of windows 7, we already have the release version from our MAPS subscription.

An employee of ours set up his gmail account to send emails as [email protected], which was fine, but now he has left the company and we have disabled his company account.

He is still able to send emails from his gmail account that appear to the recipient as from ourcompany.com. Is there any way I can block this?