morganw

If you remove the drivers, mark them as safe by pressing the unblock, then re-import - this should fix it. It is the issue listed above, you just have to be careful importing files that you've acquired from the internet. http://sonotaco.com/soft/images/unblock2.gif

You can't use chown over FTP, you would have to set ownership and permissions through some other method. If it was a server you would do it through the shell, I would have thought your NAS box would have some kind of web interface to let you configure things like this. If your NAS supports iSCSI you could use this instead and it should be easier to configure the permissions as it's seen as local storage by the server. Or getting more complicated, you could virtualise the FOG server and use NFS for the VM storage. To be honest though you could use any old workstation as a storage node if the hard disk is big enough, just queue unicast deployments and everything will still be pretty quick.

I'm pretty sure there is a .ini file somewhere on the client that you can alter instead of having to recompile anything.

I'm pretty sure that you can adjust this in a config file on the client, but I don't have a client to check. Unless you only have a few computers I wouldn't think that a really number like 10 seconds is a good idea, I imagine you could start to overload the webserver and have a negative effect on network performance.

On your fog server you can look at /etc/passwd to see the user ID: Understanding /etc/passwd File Format Hopefully on your NAS box you can set the ownership of the relevant directories with this ID number.

I think you would have to look for a umask setting on the NAS box, or look to try and match the owners id number with the id number for the FOG user on the server.

Does you VM time out requesting a DHCP address or do you get a TFTP error?

Just put C:\Program Files\SIMS.net\SIMS\Pulsar.exe into notepad, save it as sims.cmd, upload it somewhere, then link to that. It does seem a pretty strange way to start a locally installed program. Lot's of opportunity for this to clash with security / group policy restrictions and the link will be available even when SIMS isn't, so people could get confused and think they have remote access to it. I think whoever came up with the idea needs to have another think about it.

The FOG specific services are for task scheduling, image replication and multicast deployment. You won't actually need those to be running to get a machine to display the PXE menu, only the TFTP related stuff needs to be running. I think the problem is more likely to be with how you are running your DHCP server and how your network is configured in Virtual Box. If you look at the logs on your DHCP server, can you see that DHCP requests are being received? The FOG installation scripts aren't particularly robust (e.g. if you are using the newest version of Ubuntu then TFTP services won't work because one of the paths is set differently in newer packages) so I would recommend running the FOG server as a VM (so you can snapshot it during setup and upgrades), then having a physical storage node to actually serve the images from. Also, you don't need the full desktop system on the FOG server as all of the configuration is either terminal based or using the web interface.

An easy option. Reset Student User Password Free Tool for Schools, Business and Education - Burconix

Well if the Exchange server is new, and you can maintain replicas / scheduled exports / DPM backups of SIMS and the finance app then I would say local storage is more suitable. You can always add a SAN at a later date and migrate servers to it as your current equipment ages.

What are the server roles you are looking to virtualise and are they all critical? If you do the SAN route you can have full failover for everything. If you just need you domain to be available and you have a tested disaster recovery option for your storage then maybe this is good enough and your Hyper-V on SMB shares idea would be much better value for money.

You need to configure apt.conf (assuming some Debian derivative) with you proxy details so the fog install script can download packages, and then once installed, put you proxy details into FOG. As long as your SWGFL proxy will allow access to your chosen Linux distribution's repositories you won't have any problems. If you want to run a WDS and FOG mix then the easiest way is using DHCP groups and modifying the boot options per group.

I've previously used Group Policy Preferences to customise the default start menu on each workstation, and more recently redirected the start menu to a folder on each workstation's system disk (shortcuts would get copied into the folder as part of application deployment and you can restrict access with Access Based Enumeration on the share). The first way adds to Group Policy processing time but is more flexible, the second way was a lot faster but may restrict how you can deploy applications.

This may be appropriate: HYPER-V and SYSTEM CENTER: Why You Should Not Running Domain Controller in a Cluster Hyper V Host?

Spiceworks will do it, but I would recommend Lansweeper as it's easier and faster for what you want to do. Just install as a trial/freeware version and you'll have everything inventoried within a few days.

Have you tried DOMAIN\username when authenticating, rather than just typing the username?

This is the one that I used: Reset Student User Password Free Tool for Schools, Business and Education - Burconix It will run from a share and has some nice options.

How can you actually tell if Kerberos authentication is being used? I've got a valid krb5.conf (tested with kinit on the server) and I've joined the domain using the ads option, but I haven't configured a keytab for Samba. So am I right in thinking that even though the system itself can use Kerberos correctly, when I'm connecting from a client I'm actually using some other auth type like lanman or ntlm? I've only found one guide which referenced the keytab and that says to use the option "kerberos method = system keytab" in smb.conf. All the other guides I've found seem to just get the realm details into krb5.conf and then just use any old auth method when connecting. To test I've tried putting bad configuration data into krb5.conf and everything keeps working (except the kinit test).

Actually I've just found that if I tried it without the above settings it was still working. I've also just tried forcing NTLMv2 on the Windows client and client ntlmv2 = yes (I think this only allows NTLMv2 auth and nothing else) on the server, and everything is still working.

I was testing this setup last week. This was purely a test environment rather than a live system, a brand new test domain on a 2008R2 DC, and two test Samba servers with the default packages in Debian Squeeze. I tested a server with security = domain, and the other with security = ads and I seemed to get single sign-on with both approaches (tested from a Windows 7 client). I had this set at the top of the OU structure (alongside the default domain policy): Computer Configuration / Policies / Windows Settings / Security Settings / Local Policies/Security Options / Network Security / Network Security: LAN Manager authentication level - Send LM & NTLM - use NTLMv2 session security if negotiated. Is this the same as what you had done to make it work?

Out of interest, who mandates that you must have anti-virus software on your Macs?

It's like XBox Live glued onto Windows 7. I really wasn't expecting it to be so hard to use and disjointed.

Just wait a while and get the new OS X. The viruses won't run unless you buy them from the AppStore.

Is that documentation generated specifically for Debian, or does the actual Xen documentation relate nicely to the versions of Xen that packaged for Debian?