First, you should know how works FOG when you create a multicast task and the different elements involved :

[B]In the webUI:[/B]

What happend when you make click in the “download multicast tasks” button:
The php script create a “Image Package Multicast”: the script make a query to know how much computers are in the group, and create a multicast task entry in the multicastSession table (create the multicast session ID, the session port and the initial state:0 or Queue). Then, creates as many tasks as computers are in the group, and create their differents “image Package”. Those “image Package” create the PXE files and the differents registry in the tasks table. The initial state for thos tasks is 0 or Queued. And send the WOL package to the clients.

Well, we have the PXE files and the entries in the database; but what happend in the server?

[B]In the server:[/B]

We have a daemon which controls the multicast tasks, concretely we have four differents daemons, the FOGMulticastManager. This daemon queries each 10 seconds if there is a new multicast tasks. If the daemon find one, makes a query to know the session ID, how much computers and which are the image files. Then create a UPDCast command like this:

[CODE]CMD: /usr/local/sbin/udp-sender --file “/images/my_image_folder/d1p1.img” --min-receivers 20 --portbase 63142 --interface eth0 --max-wait 300 --half-duplex --ttl 32 --nokbd --mcast-data-address 239.0.104.22 --start-timeout 3600;[/CODE]

Then make an update in the database to write the new state of the multicast session. Multicast session state is 1 or In Progress. And begin to log the multicast activity in /opt/fog/log/multicast.log.udpcast.NUMBER_SESSION. The sesion have two timeouts: --max-wait 300 and --start-timeout 3600. The second one indicates how long will be activate a session before the server deletes it. If any hosts is turned on during this hour, the server will kill the session.

The min-receivers parameter indicates that the server will be waiting to 20 clients; when this magic number is reached, the session will begin, and any other computer comes after, will be out of the session (in the famous blue screen “Please Wait”).

What happend if you send a multicast session to 20 clients, and only 17 will be turn on? In this situation, the multicast session will be waiting 300 seconds, max-wait parameter, after this the session will begin for 17 clients. The other tree clients will be out of session. Remember, you have one multicast session for each image file.

When the multicast session finishes, the FOGMulticastManager, update the sesson state to 2 or Finish and kill the UDPCast process in the server.

There is the problem You can solve it easily.

When you create a task, the webUI calls to tasks.confirm.include.php ( you can locete it in WEBROOT_DIRECTORY/fog/management/includes/ ). This php file creates the PXE file depending the tasks type. For example, to create a multicast download task, the tasks.confirm.include.php calls to :

[PHP]$taskid = createImagePackageMulticast($conn, $imageMembers[$i], $taskName, $port, $tmp, (!empty($_GET[“debug”]) && $_GET[“debug”] == “true” ), true, $shutdown, $imageMembers[$i]->getKernel(), $other , (empty($bashscript)? “”:$bashscript), $type, $taskPCT);[/PHP]

This function uses the [B][I]$other[/I][/B] variable to setup the hostname, check if this variable has setup the hostname.

[quote=“Lee Rowlett, post: 12364, member: 28”]this didn’t work hostname is blank[/quote]

What is the reason to be blank? The PXE file hasn’t the hostname, the fog script not get the $hostname or the $_GET variable is empty?

You don’t need create a new query to know the hostname. Probably, the fog script knows the hostname, it is necesary to change it when you download an image. Check the PXE file to know it.

Then, you can change the web service call:

[CODE]wget -q -O /tmp/co.txt “http://${web}service/Post_Stage3.php?mac=$mac&imgid=$imgid&hostname=$hostname” &>/dev/null[/CODE]

Capture the hostname in the Post_Stage3.php file:
[PHP]$hostname = $_GET[“hostname”];[/PHP]

I think that develope it is really easy.

There are tree web service (they are located in WEBROOT_DIRECTORY/fog/service):

[LIST]
[]Pre_Stage1.php . This web service is called when a task begins.
[]Post_Stage2.php . This web service is called when an upload task ends .
[*]Post_Stage3.php . This web service is called when an download task ends .
[/LIST]

Those web services are called by fog script either when you upload either when you download an image. Those web services responde with an echo (##), if the proccess finishes well. You can call to the [I][B]mail[/B][/I] php function (you can see the sintaxis [URL=‘http://php.net/manual/en/function.mail.php’]here*[/URL] ) to send an email. You must call it before the echo.

===============================

• [url]http://php.net/manual/en/function.mail.php[/url]

Clamav has a new version, 0.97.7; then, is probably that our version is obsolete. I will explain how to update it.
You must use the different files that are in the clamav0.97.7.zip file. It contains next files:
[LIST]
[]clamav-config
[]clambc
[]clamconf
[]clamdscan
[]clamscan
[]fog.av
[]freshclam
[]libclamav.so.6.1.16
[/LIST]
==> [B]Updating our server:[/B]

You can see some information in this [URL=‘http://www.fogproject.org/wiki/index.php/Clamav’]link[/URL][B] [1][/B]
[LIST]
[*]Updating clamav version:
[/LIST]
For RHEL 6:
[CODE]$ sudo yum install clamav[/CODE]
For Ubuntu:
[CODE]$apt-get install clamav-freshclam[/CODE]

[LIST]
[]Deleting the old definitions files in the server:
[/LIST]
[CODE]$ sudo rm /var/lib/clamav/.*[/CODE]

[LIST]
[]Update your definitions by running freshclam
[/LIST]
[CODE]$sudo freshclam[/CODE]
[LIST]
[]Disable scripted updates and let ClamAV download definitions to the root of the web server. Add “ScriptedUpdates off” to the freshclam.conf file on the Fog Server.
[/LIST]
==> [B]Updating init.gz[/B]

[LIST]
[*]Expand the Fog init.gz file:
[/LIST]
[CODE]
$sudo cd /tftpboot/fog/images
$sudo gunzip init.gz
$sudo mkdir initmountdir
$sudo mount -o loop init initmountdir
[/CODE]

[LIST]
[*]Extract clamav0.97.7.zip files (you have copied it in your server home folder) and copy them:
[/LIST]
[CODE]$cd
$sudo unzip clamav0.97.7.zip
$sudo cp clamav-config /tftpboot/fog/images/initmount/usr/bin
$sudo cp clambc /tftpboot/fog/images/initmount/usr/bin
$sudo cp clamconf /tftpboot/fog/images/initmount/usr/bin
$sudo cp clamdscan /tftpboot/fog/images/initmount/usr/bin
$sudo cp clamscan /tftpboot/fog/images/initmount/usr/bin
$sudo cp freshclam /tftpboot/fog/images/initmount/usr/bin
$sudo cp libclamav.so.6.1.16 /tftpboot/fog/images/initmount/usr/lib
$sudo cp fog.av /tftpboot/fog/images/initmount/bin
[/CODE]

[LIST]
[*]Create the soft links:
[/LIST]
[CODE]$cd /tftpboot/fog/images/initmount/usr/lib
$ln -s libclamav.so.6.1.16 libclamav.so
$ln -s libclamav.so.6.1.16 libclamav.so.6
$chmod 755 libclamav.so.6.1.16[/CODE]

[LIST]
[]Modify the freshclam.conf file inside the init. Add “DatabaseMirror MyFogServerNameOrIP” and “ScriptedUpdates off” to this file.
[/LIST]
[LIST]
[]Change the ram disk size. The old size is 50 MB, is small, then we will increase it to 100 MB:
[/LIST]
[CODE]$sudo vi /tftpboot/fog/images/initmount/bin/fog.av[/CODE]
We edit the next line:
[CODE]mount -t tmpfs none /usr/share/clamav/ -o size=50m;[/CODE]
And change the ram disk size to 100m:
[CODE]mount -t tmpfs none /usr/share/clamav/ -o size=100m;[/CODE]
Save the change

[LIST]
[*]Compress the init file:
[/LIST]
[CODE]$cd /tftpboot/fog/images
$sudo umount initmountdir/
$sudo rmdir initmountdir
$sudo gzip init[/CODE]

[LIST]
[*]DONE the update!!!
[/LIST]

[1] [url]http://www.fogproject.org/wiki/index.php/Clamav[/CODE][/url]

[url=“/_imported_xf_attachments/0/296_clamav0.97.7.zip?:”]clamav0.97.7.zip[/url]

I have found a little bug.
In Inventory.php (management/reports/Inventory.php) the SQL has a mistake: the hostOS atribute not exists in hosts table.
[CODE]— Inventory.php (revisión: 1)
+++ Inventory.php (revisión: 2)
@@ -73,8 +73,7 @@
FROM
hosts
inner join inventory on ( hosts.hostID = inventory.iHostID )

•          left outer join images on ( hostImage = imageID )
  

•          left outer join supportedOS on ( hostOS = osID )";
  

•          left outer join images on ( hostImage = imageID )";
  

$res = mysql_query( $sql, $conn ) or die( mysql_error() );

while ( $ar = mysql_fetch_array( $res ) )
[/CODE]

If you run the SQL in DB:

[I]mysql> SELECT [/I]
[I] -> * [/I]
[I] -> FROM [/I]
[I] -> hosts [/I]
[I] -> inner join inventory on ( hosts.hostID = inventory.iHostID )[/I]
[I] -> left outer join images on ( hostImage = imageID )[/I]
[I] -> left outer join supportedOS on ( hostOS = osID );[/I]
[I]ERROR 1054 (42S22): Unknown column ‘hostOS’ in ‘on clause’[/I]

Hi,

We have developed a feature to make a software inventory. Although we use “FOG 0.30” (it’s a chimera between 0.30, 0.32 and 0.33 :)) I think is possible to implement quickly into FOG 0.33.

The feature works in this way:

1º) When you upload an image, the fog script import some windows registry keys ( only on Windows XP and W7, i don’t test it in W8).
2º) Before the end of the upload process, a new web service process the different registry files and insert the program data in DB (Program Name, Version, uninstall Path, …) and link this inventory to the image ID.
3º) The web service insert into History table a little log: which image Name and ID, from which computer and software inventory Version. When you upload a new image version, a new sofware inventory version is created.
Example: Inventario de software ( [B]ver 18[/B] ) añadido a la imagen [B]aula-bz-ingenieros-p0b14-docencia[/B] desde [B]u001022[/B]
You can see that is the 18º version, image name "aula-bz-ingenieros-p0b14-docencia[B] " [/B]and from computer name u001022
4º) when you download a image to one computer, the web service (Post_stage3.php) logs in DB which software version is downloading.
Example: Imagen aula-bz-rectorado-cpd-windows7Duado (version 1) volcada en equipo u001022
Image name “aula-bz-rectorado-cpd-windows7Duado”, version 1 and to u001022
5º) You can consult the diffent software versions by one report.

[ATTACH]279[/ATTACH][ATTACH]280[/ATTACH][ATTACH]281[/ATTACH]

[url=“/_imported_xf_attachments/0/278_FOGReport.pdf?:”]FOGReport.pdf[/url][url=“/_imported_xf_attachments/0/279_ConsultingInventoryVersions.png?:”]ConsultingInventoryVersions.png[/url][url=“/_imported_xf_attachments/0/280_ImagesoftwareInventory.png?:”]ImagesoftwareInventory.png[/url][url=“/_imported_xf_attachments/0/281_SoftwareInventoryReport.png?:”]SoftwareInventoryReport.png[/url]

I’am developing new features to use FOG in my university, and I think that those features would be interesant to FOG comunity, but I don’t know the channels to contribute with the project. How can I do it?

Hi, syschuck.
The code that you can find in [URL=‘http://www.ehu.es/fogenehu’]www.ehu.es/fogenehu [/URL]is actually in Production, and is the code that we are using now. But is obsolete, is based on FOG 0.30, and have a limited support to W7. You can use it to see how is developed the different users levels.

User.class.php
[PHP]
class [URL=‘http://www.ehu.es/fogenehu/doc/html/class_user.html’]User[/URL]
{
private $strUserName, $strAuthIP, $strTime, $strType, $strPass, $strLanguage;
private $intID, $strIdentity;

const TYPE_MOBILE = ‘0’;
const TYPE_ADMIN = ‘1’;
const TYPE_GESTOR= ‘2’;

function __construct($id, $username, $authIp, $authTime, $type, $identity)
{
$this->intID = $id;
$this->strUserName = $username;
$this->strAuthIP = $authIp;
$this->strTime = $authTime;
$this->strType = $type;
$this->strPass = null;
$this->strIdentity = $identity;
$this->strLanguage = null;
}

And more…[/PHP]

During login process [See management/includes/processlogin.include.php code*], you obtain a User object with the user rol [MOBILE, ADMIN and the new one GESTOR]. With this value you can filter the access, for example: We filter the access to some menus, the ADMIN rol can see the storage node configuration or Profile menu but a GESTOR rol can’t access to it [See management/include/mainmenu.include.php].

[CODE]
<a href=“?node=” class=“home menu_button” alt=“<?php echo((“home”)); ?>" title="<?php echo((“home”)); ?>”></a>
<a href=“?node=users” class=“users menu_button” alt=“<?php echo((“users”)); ?>" title="<?php echo((“users”)); ?>”></a>
<?php
if ( $currentUser->getType() == User::TYPE_GESTOR )
echo ( “<a href=‘?node=profiles&sub=list’ class=‘profiles menu_button’ alt=‘centros’ title=‘centros’></a>” );
?>
<a href=“?node=host” class=“host menu_button” alt=“<?php echo((“hosts”)); ?>" title="<?php echo((“hosts”)); ?>”></a>
<a href=“?node=group” class=“group menu_button” alt=“<?php echo((“group”)); ?>" title="<?php echo((“group”)); ?>”></a>
<a href=“?node=images” class=“images menu_button” alt=“<?php echo((“images”)); ?>" title="<?php echo((“images”)); ?>” ></a>
<?php
if ( $currentUser->getType() == User::TYPE_GESTOR )
echo ( “<a href=”?node=storage" class=“storage menu_button” alt=‘“.(“storage”).“’ title='”.(“storage”).”’ ></a>" );
?>[/CODE]

[*] You can see in the code the ldap validation support

The changes are a lot of, but you need:

*Create some tables in DB: Profiles and ProfileMembers. Those tables contain the infomation about the Profile (profile as departament or building, not as a rol) and the relationship between the users and those Profiles.
*Add a new atribute in Group table: the profileID.

The relationships are those:
Users <–> ProfileMembers <–> Profiles <–> Groups <–> GroupMembers <–> Hosts

Our image names have a pattern: groupName-service
Where service is or “docencia” or “otros”. For example: aula-bz-ciencias-AI1-docencia, the group name is aula-bz-ciencias-AI1 and the service, docencia.
Thus when you need filter the access to one image you use the group name:
Select * from images where imageName like ‘Your_Group_Name’

We do in this way to not make a new relationship between Profiles and Images:
Users <–> ProfileMembers <–> Profiles <–>ImagesAssoc <–> Images

• Add a new atribute in the users table. You need the user rol : mobile, technician and admin.
• Modificate the User Class to add this rols

You can see info in [URL=‘http://www.ehu.es/fogenehu’]www.ehu.es/fogenehu[/URL] and the code is in [url]http://www.ehu.es/fogenehu/doc/html/index.html[/url] (this doxygen doc is actually obsolete)

[quote=“Martin T, post: 10917, member: 3420”]Thank you Fernando Gietz and chad-bisd for input.

I have rebuilt the complete fog server with Ubuntu 12.04.2 LTS 64bit and fog .32 and I am back to the same error that I started with (clamaverror.jpg Monday 2:41PM). I did this because I have tried so many changes that I wanted to make sure that I got back to a safe starting point. So a few questions, if you please.
Do I go back and make all the changes 1-9, yesterday at 11:34 AM or try the rebuild/resize of init.gz?[B] - both[/B]
The server does have 0.97.6 and I see the client does say 0.97.1. Are the definitions truly incompatible?[B] - Fernando Gietz[/B]
Where is the FOG_PXE_IMAGE_DNSADDRESS at? [B]- Fernando Gietz[/B]
I can only find main, daily, and bytecode CVD files. No CLD files. Do I just make the symbolic links to what I have? [B]- chad-bisd[/B]
I have asked other questions above but let’s see how this goes.
I am willing to try almost anything and can rebuild the server when I know exactly the steps to make it work. Hopefully I can create a complete posting that outlines the exact steps to make this work for everyone who wants this task to work correctly.[/quote]

Sorry for answer so late.

[I]Where is the FOG_PXE_IMAGE_DNSADDRESS at?[/I]
In webui: About > FOG settings > TFTP Server

[I]The server does have 0.97.6 and I see the client does say 0.97.1. Are the definitions truly incompatible?[/I]
I’m not claAV expert but the capture says that [I]Not supported database files founf in /usr/share/clamav. [/I]Seems the database files are incompatibles. This problem/issue desappears if you upgrade the clamAV version[I] to 0.97.6[/I]

I have seen the fog.av script and you don’t must resize the init.gz (sorry, but i’m sure that you have learn a lot of ). The database files from the server are copied in /usr/share/clamav. This folder is a ramdisk:

[CODE]mount -t tmpfs none /usr/share/clamav/ -o size=50m;[/CODE]

This ramdisk have 50M size. This size would be little[I], why? [/I]the reason is the database files size grown more and more, daily. For example, in my server those files:

[CODE]-rw-r–r-- 1 clam clam 302K feb 15 03:37 bytecode.cld
-rw-r–r-- 1 clam clam 0 feb 26 20:25 clamd.sock
-rw-r–r-- 1 clam clam 55M mar 12 03:10 daily.cld
-rw-r–r-- 1 clam clam 30M mar 11 14:14 main.cvd
-rw-r–r-- 1 clam clam 572 mar 12 03:10 mirrors.dat[/CODE]

86 MB, and daily increases his size The size of the ramdisk would be dinamic.

You must config the FOG_PXE_IMAGE_DNSADDRESS value, if you don’t have configured it, the client doesn’t know how to resolve the direction of the clamAV server.
To rebuit the init.gz with the new version of clamav
[url]http://fogproject.org/forum/threads/clamav-version-update.3926/[/url]
To resize the init.gz
[url]http://fogproject.org/forum/threads/change-initrd-size.3920/[/url]

The antivirus in FOG works as follow:

[B]In the server:[/B]
During the FOG install process, you install clamav package from the repositories. Actually the clamAV version is 0.97.6.
the freshclam runs every day, you can setup it in freshclam.conf file, and download the new virus definitions and update the files: bytecode.cvd and daily.cvd.

[B]In the client:[/B]
In the init.gz the clamav is installed, you can see his version in the screen (your capture says that the clamAV version is 0.97.1). Opss, your version isn’t the same in the server and in the client. The server version is 0.97.6 and the client 0.97.1, the definitions are differents.
The script fog.av try to download the virus definitions from your FOG server, but the script says “Not supported database files”. The reason of this is clear, the server version files (0.97.6) and the client version files (0.97.1) are differents, and the definitions are not compatibles. Well, is one little problem, but is not the world end. The script, then, tries to connect to the clamav servers, you must setup it in /etc/freshclam.conf file (step 6,7,8), and downloads the virus definitions for your clamav version (0.97.1). If the download ends, the antivirus runs the scan. Be sure that you have setup the DNS_ADDRESS in FOG server settings

Well, install the clamav in the server, i think, is not necessary. His mission is “only” download the virus definitions every day.

There are some little problems with this feature:
1º) The init.gz free space is little and the definitions files are more and more large, and don’t fit in the free space.
2º) You must have the same version in the server and in the client. To do it, you must rebuilt the init.gz

I have the same problem, but finally I resolve it.
If you open, in the browser, the next file:
[url]http://Direction_of_your_FOG_server/fog/status/bandwitch.php[/url]
The server must response an echo. If the server not have activity, the response must be 0##0. In my case, the response was nothing. At the begining I thougth the problem was the cURL library (I have problem with it in the past), but after some test, I discarded it. Finally I found the answer to my problem: PHP tags.

The server did not interpret well the php short tags, <? ?>. If you change it for <?php ?>, works fine. If you don’t want make code changes, you can also configure the php.ini file to support short open tags.

#vim /etc/php.ini
Edit the line:
short_open_tag = On

Save the changes
#service httpd restart

PD: I use RHEL 6.3, but you have the same problem in Centos 6

[IMG]http://www.ehu.es/fogenehu/captures/imagenes/FOG_usersMenu.jpg[/IMG]
The solution, maybe, is create a new management layer, and not create a new atribute in Host class.
See, we have modificated FOG code to add a new layer call “Site”. The hosts are linked to one “Site” and one or more technicians are linked to the “Site”, too. The technician only see the hosts and the images of his “Site”.

[url=“/_imported_xf_attachments/0/242_FOG_usersMenu.jpg?:”]FOG_usersMenu.jpg[/url]

Well I have install a FOG 0.30 in RHEL 6.3, and you must make some changes:

1º) Disable firewall and selinux
[I]#sudo selinuxenabled; echo $?[/I]
if the echo returns 0, the SELinux is enable and you must disable it
[I]#sudo /etc/selinux/confi[/I]g
change the value to disabled
[I]#sudo setenforce 0[/I]
[I]#reboot[/I]
2º) Download the FOG code
3º) You must make some change in some config files:
3.1) Unziped_folder/lib/redhat/config.sh
[I]packages=“httpd [B]php php-gd php-cli[/B] mysql mysql-server [B]php-mysql[/B] tftp-server nfs-utils vsftpd net-tools wget xinetd tar gzip make m4 gcc gcc-c++ htmldoc perl perl-Crypt-PasswdMD5 lftp clamav [B]clamd[/B]”;[/I]
I change the next packages: php53 -> php and add clamd. The php-gettext not exists in RHEL.
4º) download and install the next repositories:
[I]#wget [url]http://apt.sw.be/redhat/el6/en/x86_64/rpmforge/RPMS/rpmforge-release-0.5.2-1.el6.rf.x86_64.rpm[/url][/I]
[I]#wget [url]http://mirror.uv.es/mirror/fedora-epel/6/i386/epel-release-6-8.noarch.rpm[/url][/I]
[I]#sudo rpm -i --quiet rpmforge-release-0.5.2-1.el6.rf.x86_64.rpm[/I]
[I]#sudo rpm -i --quiet epel-release-6-8.noarch.rpm[/I]
The EPEL repository is necesary to install htmldoc package.
5º) Run installfog.sh to install FOG
6º) You must config timezone in your php.ini:
[I]#sudo vim /etc/php.ini[/I]
In my case:
date.timezone = “Europe/Madrid”
[I]#/etc/init.d/httpd restart[/I]

Some issues:
*During the install process appears the next error:
[I]Setting up and starting NFS Server…Failed![/I]
Solution:
[I]#sudo service rpcbind start[/I]
*I have config the logwatch to see the freshclam messages, to config it:
Edit the [I]/etc/logwatch/conf/logwatch.conf[/I] file
Add the next line:
[I]MailTo = [EMAIL]miemailaddress@mydomain.com[/EMAIL][/I] (<== your email)
if you don’t want the logwatch messages delete the [I]/etc/cron.daily/0logwatch[/I] file

• In logwatch messages appears the ext error:
  [I]Notifyclam error[/I]
  Solution:
  Edit the [I]/etc/freshclam.conf[/I], uncomment and config the next line:
  [I]NotifyClamd /etc/clamd.conf[/I]
  [I] Error: clamd was not notified can’t connect to clamd through /var/run/clamav/clamd.sock[/I]
  Solution:

sudo touch /var/lib/clamav/clamd.sock

sudo chown clam:clam /var/lib/clamav/clamd.sock

Well, at last I make it I have clamav 0.97.6 installed in my init.gz.
I have remade my init.gz usinig builtroot (see[SIZE=3] [URL=‘http://www.fogproject.org/wiki/index.php/Build_FOG_file_system_with_BuildRoot_-_init.gz’]Build FOG file system with BuildRoot - init.gz[/URL][/SIZE] [1])
But you need make one change in file …/buildroot-2011.08/package/customize/clamav/clamav.mk
You must change [B]CLAMAV_VERSION=0.97.6[/B]

[CODE]#############################################################

clamav

#############################################################
CLAMAV_VERSION:=0.97.6
CLAMAV_SOURCE:=clamav-$(CLAMAV_VERSION).tar.gz
CLAMAV_SITE:=http://downloads.sourceforge.net/project/clamav/clamav/$(CLAMAV_VERSION)
CLAMAV_INSTALL_STAGING=YES
CLAMAV_LIBTOOL_PATCH=NO
CLAMAV_CONF_OPT= --disable-clamav
–enable-clamdtop=no
–disable-clamuko
–enable-check=no
–enable-llvm=no
–program-transform-name=
CLAMAV_DEPENDENCIES=zlib bzip2

define CLAMAV_REMOVE_DB
rm -f $(TARGET_DIR)/usr/share/clamav/*.cvd
endef

define CLAMAV_SIMPLE_CONFIG
echo “DatabaseMirror database.clamav.net” > $(TARGET_DIR)/etc/freshclam.conf
echo “DatabaseOwner root” >> $(TARGET_DIR)/etc/freshclam.conf
endef

CLAMAV_POST_INSTALL_TARGET_HOOKS += CLAMAV_REMOVE_DB
CLAMAV_POST_INSTALL_TARGET_HOOKS += CLAMAV_SIMPLE_CONFIG

$(eval $(call AUTOTARGETS,package/customize,clamav))[/CODE]

Also you have erase the “/” in line:

[QUOTE]Old line:
CLAMAV_SITE:=[url]http://downloads.sourceforge.net/project/clamav/clamav/$(CLAMAV_VERSION)/[/url]
New line:
CLAMAV_SITE:=[url]http://downloads.sourceforge.net/project/clamav/clamav/$(CLAMAV_VERSION)[/url]
[/QUOTE]

[1] [url]http://www.fogproject.org/wiki/index.php/Build_FOG_file_system_with_BuildRoot_-_init.gz[/url]

And FOG 0.33 beta can resize W7 sysprep monopartition images?

As monopartition images in W7, we understand a W7 image with: 100 MB W7 special partition + XXX GB W7 system partition, no?