Posts made by Bob Henderson

Would run once work with domain join/rename bit of Fog? I’m worried that with the fact it usually gets to a login before domain join, then reboots, it would run and then fail.

[quote=“Wayne Workman, post: 47235, member: 28155”]If you have a domain, there is zero reasons to “Deploy” a powershell script.

Powershell can control domain-connected clients like puppets. You can literally push out whatever you want to whatever computer you want, instantly, from powershell itself on a DC.[/quote]

Completely understood. Problem is, I’m doing something a bit different. To lay it out easily:

1. I create a base image (Windows 7 + Office2013) and get it all sysprepped and into FOG.
2. I deploy base image to various hardware.
3. I have this script to invoke a deployment job using PDQ deploy to install the rest of my packages in their updated versions.

I’d love to get it so that that script fires off automatically. Problem is, since it’s powershell using WinRM to invoke the PDQ Deploy command, I need to be domain joined. If I’m not mistaken, this puts adding the script to the SetupComplete.cmd file out of the question, as the domainjoin from FOG hasn’t ran yet.

Any suggestions?

[quote=“Bob Henderson, post: 47213, member: 1291”]Any more updates at this? I’ve just got the script working to pull updates from PDQDeploy on the machine, but am wondering how to fire it off after an image. Are you still using a snappin, or something different? I’d love to get this r unning, as I’ve got 1200 machines to image next month…[/quote]

And yes, thread necro, but for relevance!

[quote=“BigMan99211, post: 24135, member: 21932”]OOOHH… Just had an epiphany as 5am…

Better yet, I’ll just have a machine group set up in PDQ Deploy, with all of the same machines that are in the FOG group. This way, I can just have the script talk directly to the PDQ Deploy machine group, instead of going all the way out and through FOG.[/quote]

Any more updates at this? I’ve just got the script working to pull updates from PDQDeploy on the machine, but am wondering how to fire it off after an image. Are you still using a snappin, or something different? I’d love to get this r unning, as I’ve got 1200 machines to image next month…

[quote=“Wayne Workman, post: 45752, member: 28155”]IMHO, just add a hard drive to FOG. Use the cloud backup as a backup…[/quote]

…works fine, as long as it’s not your only backup.

Ack, didn’t even think of that. We’re still on all legacy bios machines due to their age (but they still run the 10 technical preview fine) , but that’s something to consider moving forward…

Thanks!

Ok, so I know I’m jumping the gun here, so stick with me.

With Windows 10 being released sometime this summer, how much of a hassle should it be to get FOG to work with it? I haven’t tried, and a cursory search of the forums didn’t lead to much, so I’m wondering.

Thanks!

[quote=“Wolfbane8653, post: 42008, member: 3362”]You could also run unicast in groups. Connect them all to the switch, register, add them all to “group1”, and tell the group to download.

Multicast is [B][U]not[/U][/B] the only method to do mass amounts of machines at once. I do a lab of 30 in a clip using unicast groups. Set em up and walk away for an hour. Then come back to setup the next set.

BUT Uncle Frank is right please try the wiki and post the results.

Correct wiki page is [url]http://fogproject.org/wiki/index.php?title=Multicasting[/url]
The information is the same due to a merge of the pages.[/quote]

That’s what I’m doing right now, mostly. and honestly, it works. I just hoped with Multicast it would take less time.

2 clients, since this was just a test. I tried two different ones, in a different chunk of the building on a different switch, same result.

Unicast is uniformly fast. I haven’t ran into one that is slower. We’re gig across the board, and it has helped out alot.

that was my understanding of Multicast too, either 100% or nothing. Hence the confusion.

I’ll dig through the troubleshooting. For now, I’m just Unicast and queing, but doing a whole group in one shot would be fantastic. We’re 1:1 laptops, and I’d love to be able to hook up a whole room to the switch,s tart, go for coffee, and come back.

So, trucking along with Fog. Now trying multicast.

I’ve got the one Fog server, which unicasts flawlessly. Network is flat, hooked up via numerous Netgear GSM7248v2 switches. These all have IGMP snooping and query enabled, as thats what I’ve seen recommended for multicast.

Unicast speeds are approx 3.5 GiB/min to the client.

Multicast speeds are approx 3.4 MiB/min to the clients.

I’ve replaced all network cables that can be easily replaced (switch patch cords, cords to clients, etc) with new to verify them. Did a cable check on the main links between the switches, all come up fine. All cables are Cat6 or better.

Any suggestions to try?

Perfect. I wasn’t aware that even existed. Leave a project alone for a while, come back to a snazzy new website and all my questions answered!

Bah, the formatting took out my spacing. Fog2 and Storage Node should be under {site b}

Alright, now that you guys helped with my wireless MAC option (thanks!) I’ve got another question.

I have 2 sites, both connected via a VPN. Currently, I have 2 different FOG servers, one at each site. They both host the same images, snapins, etc.

If I was to remove one machine, and replace it with just a storage node, would my clients in the site with the storage node pull from it automatically? I don’t want the clients to pull the image across the vpn for bandwidth reasons. But I would love to be able to manage one box and not duplicate effort all the time.

I think i’ve explained myself, but if not, ASCII graphic of network incoming:

Current network
{site a} ---------------VPN-----------------{site b}
[Fog1] [Fog2]

Client boots in site a, dhcp gives them fog 1, they pull from fog 1. Same in Site b

Proposed Network:

{site a} ---------------VPN-----------------{site b}
[Fog1} [StorageNode]

Client boots in either site, DHCP gives them fog 1, and if in Site b, pulls from that storage node over the lan instead of the vpn.

Completely nevermind. I was looking at host that was named incorrectly. I see it now. No mass approval, I assume?

Where would I see that? I’m looking a host record now that has been in deployment for 6+ months since it was last imaged, but don’t see an “approve” option.

Or am I looking in the completely wrong place?

Perfect. Now, since I have 900+ hosts in my inventory, I assume theres no way to pull these in automatically?

We use fog to deploy out our 1:1 laptop solution. I just discovered the joy of snapins, and would love to use them. Problem is, our users connect exclusively through wifi, which of course has a different mac address than the NIC used to image with fog.

Is there a way to allow snapins to deploy to the secondary, wifi MAC instead of only the LAN one? Not only that, but things like domain join, etc, would be fantastic if they worked over wireless. I see in the logs the machine checks in to fog fine, but it checks in using the wifi mac, and therefore sees no jobs.

I think I’ve worded this right.

That’s exactly what I’m looking for. Thanks guys!

I’ve just setup a new server for a client, and would like to transfer the host list over to it. Is this possible without having to re-register?

I’ve already transferred over the images by creating an image on the new machine, then SCP’ing the image file over into that folder so it’s all there.

Never done this before, and I apologize if I searched incorrectly and missed a topic that shows this.