Remote Accessing my Desktop from my laptop

[Aeris_X]

So heres some context, I am a college student majoring in Cyber security. I usually have extremely resource intensive tasks that I need to run on a computer and built a tower (specs at bottom if interested) in order to do so. In some cases, I need to do these tasks in case I forget to do an assignment but my laptop, unlike my tower, is horrible and can't run these tasks. Obviously due to COVID I have had full access to my desktop however, since I have the freetime I thought it would be a good idea to set this up ahead of time so if I am in a pinch I have full access to my pc. Here are somethings I want to be able to do:

- Be able to turn on the computer from boot. (What i mean here is being able to start the computer from when it is fully shut down)

- Use the full power of the computer

- Have full access to the PC including R/W and Admin perms

- Secure access

- Be able to see both of my monitors

 

If anyone can find my a solution that would be very appreciated. 

 

 

for those who are interested here are the specs:
R7 2700x RTX 2070 ti 16gb dddr4-3200


I have also tried using Chrome Remote Desktop but I really want the ability to connect and boot the pc so I can save power.

[Kisai]

1. You can not remotely turn on a remote PC at will. Wake on LAN is a "LAN" feature, thus requires the router to pass that information, which means that anyone else can also wake it. It's not secured. So WoL is basically something you would need to trigger from a router first, or have something inside the network listening for a "please wake me" using some outside polling source.

2. You can not use the full power of the PC, ever. At best

a) Remote desktop will transmit the video and audio of the PC as though a software GPU and audio device are installed. It will not use the GPU* for this.

b) Other RAS (Remote Access Software) is closer to streaming the PC as a video, at the expense of latency. So the frame rate will be higher if treated like video, but usually will be capped to about 15fps.

3. Remote Desktop is secure enough, if you want to make sure however you may need a ssh relay and go that route first.

4. Nope, that is definitely not happening from a remote. Even if the software can show you both monitors usually the result is the entire desktop is resized to the client machine's desktop instead.

 

About GPU usage: 

https://docs.microsoft.com/en-us/windows-server/remote/remote-desktop-services/rds-graphics-virtualization

[LIGISTX]

2 minutes ago, Aeris_X said:

So heres some context, I am a college student majoring in Cyber security. I usually have extremely resource intensive tasks that I need to run on a computer and built a tower (specs at bottom if interested) in order to do so. In some cases, I need to do these tasks in case I forget to do an assignment but my laptop, unlike my tower, is horrible and can't run these tasks. Obviously due to COVID I have had full access to my desktop however, since I have the freetime I thought it would be a good idea to set this up ahead of time so if I am in a pinch I have full access to my pc. Here are somethings I want to be able to do:

- Be able to turn on the computer from boot. (What i mean here is being able to start the computer from when it is fully shut down)

- Use the full power of the computer

- Have full access to the PC including R/W and Admin perms

- Secure access

- Be able to see both of my monitors

 

If anyone can find my a solution that would be very appreciated. 

 

 

for those who are interested here are the specs:
R7 2700x RTX 2070 ti 16gm dddr4-3200

For waking the PC, I doubt any consumer mobo will allow you to turn it on via ethernet, but if its just asleep, wake on lan would work. Obviously in this case, if you are outside of your LAN, you would need to open up a port for that in your router which from a security standpoint sorta sux. You could set up a raspberry pi on your LAN, open that up to SSH via RSA keys only, harden it as much as possible (which is plenty hardened to be honest, at least IMO), SSH into that, and then have the r-pi WOL your desktop. As far as remote access, I use team viewer but I am not sure that will be what you need.

[Aeris_X]

Just now, LIGISTX said:

For waking the PC, I doubt any consumer mobo will allow you to turn it on via ethernet, but if its just asleep, wake on lan would work. Obviously in this case, if you are outside of your LAN, you would need to open up a port for that in your router which from a security standpoint sorta sux. You could set up a raspberry pi on your LAN, open that up to SSH via RSA keys only, harden it as much as possible (which is plenty hardened to be honest, at least IMO), SSH into that, and then have the r-pi WOL your desktop. As far as remote access, I use team viewer but I am not sure that will be what you need.

 

Just now, Kisai said:

1. You can not remotely turn on a remote PC at will. Wake on LAN is a "LAN" feature, thus requires the router to pass that information, which means that anyone else can also wake it. It's not secured. So WoL is basically something you would need to trigger from a router first, or have something inside the network listening for a "please wake me" using some outside polling source.

2. You can not use the full power of the PC, ever. At best

a) Remote desktop will transmit the video and audio of the PC as though a software GPU and audio device are installed. It will not use the GPU* for this.

b) Other RAS (Remote Access Software) is closer to streaming the PC as a video, at the expense of latency. So the frame rate will be higher if treated like video, but usually will be capped to about 15fps.

3. Remote Desktop is secure enough, if you want to make sure however you may need a ssh relay and go that route first.

4. Nope, that is definitely not happening from a remote. Even if the software can show you both monitors usually the result is the entire desktop is resized to the client machine's desktop instead.

 

About GPU usage: 

https://docs.microsoft.com/en-us/windows-server/remote/remote-desktop-services/rds-graphics-virtualization

 

 

 

Interesting, thanks for the information!
 

 

So I have seen my dad from when i was younger use his desktop from work at home on his laptop using an RSA token and some VM program, would this be realistically viable for what I want to use it for? ( Dont know how he did it)

[LIGISTX]

1 minute ago, Aeris_X said:

 

 

 

 

Interesting, thanks for the information!
 

 

So I have seen my dad from when i was younger use his desktop from work at home on his laptop using an RSA token and some VM program, would this be realistically viable for what I want to use it for? ( Dont know how he did it)

I think it just very depends what your doing.... Windows has linux subsystem now, so you could just SSH directly into your windows box, but you wouldn't be able to run any windows programs, and there won't be a GUI, it will obvi just be SSH. 

 

I don't know of a way to get into windows cuz RSA keys, and I am not sure what you mean by some VM program. No real reason to virtualize anything here, unless you have a specific reason for virtualization?

[Aeris_X]

Just now, LIGISTX said:

I think it just very depends what your doing.... Windows has linux subsystem now, so you could just SSH directly into your windows box, but you wouldn't be able to run any windows programs, and there won't be a GUI, it will obvi just be SSH. 

 

I don't know of a way to get into windows cuz RSA keys, and I am not sure what you mean by some VM program. No real reason to virtualize anything here, unless you have a specific reason for virtualization?

The only reason I mention virtualization is because Ive seen my dad use a VM program and an RSA token to access a desktop at work with pretty mcuh everything I want to do with my pc. It was setup by his workplace so he doesnt know anything about it and ive been on a hunt to find out how it works

[Rolling2405]

13 minutes ago, Kisai said:

Wake on LAN is a "LAN" feature

II am very interested in learning how to do this please post a guide or tutorial on how to do this using Teamviewer or something else free

[LIGISTX]

2 hours ago, Aeris_X said:

The only reason I mention virtualization is because Ive seen my dad use a VM program and an RSA token to access a desktop at work with pretty mcuh everything I want to do with my pc. It was setup by his workplace so he doesnt know anything about it and ive been on a hunt to find out how it works

I mean, is he logging in through a web browser, using a token to authenticate him and then the server spins up a VM for him to work in?

 

If so, that is likely some vmware software, but that likely is not what you want. VM's are useful for corporations to slice servers down so more things, or people, can use little slices at once. You are your computers only user, no reason to do anything with VM's unless you want to run multiple different dev environments at the same time.

[Kisai]

3 hours ago, Aeris_X said:

 

 

 

 

Interesting, thanks for the information!
 

 

So I have seen my dad from when i was younger use his desktop from work at home on his laptop using an RSA token and some VM program, would this be realistically viable for what I want to use it for? ( Dont know how he did it)

RDP and Citrix clients behave in a similar way. Like ever since Windows 2000 "Terminal Services" have been available on all versions of Windows, which means that a desktop can access another desktop or a server running Windows 2000 or later (or NT4 that had terminal server on it) 

 

The gotcha in this regard, is that terminal services on a desktop is a single-session environment, so even though yes, the machine does support suspending multiple users on the machine, only one user actually has control over the machine, so if you login remotely, the user on the local end is booted off. 

 

Things like Teamviewer, actually run in user-space, so you have to actually login to the machine before it's available. Likewise for VNC.

 

A thing I used to do with my previous generation of pc's and mac's was to control the mac from the pc or the pc from the mac using a remote "kvm" type of tool, and then flip the screen input to the actual machine. So any time the mouse moved to the "remote" desktop, I was actually looking at the real desktop. It has the same drawbacks as a virtual machine though, as the input is not 1:1 if the monitor is not the same resolution, and applications on the remote system can only deal with relative input, so things like games often just flip out and have no precision to the movement. You're better off leaving a game controller plugged in.

 

[Kisai]

4 hours ago, Thomas001 said:

II am very interested in learning how to do this please post a guide or tutorial on how to do this using Teamviewer or something else free

OK so for wake on lan, you have to enable it in the Bios, and then you need to enable it in Windows in the network driver. Intel drivers have this support, Realtek is a bit of a crapshoot, and likely USB network adapters can't support it since the power will be turned off to the USB device unless a power share is turned on in the bios as well.

 

On linux (eg if you use an rPi or something) there is a tool called etherwake. Enter the MAC address of the machine you want to wake up, and that's it. The machine will either come out of suspend, or come out of a power-off state as long as the PSU power is still turned on. If the physical switch is flipped, then of course it won't turn on.

 

There's an example of how to do this here:

https://notenoughtech.com/featured/use-raspberry-pi-wol/

 

If you do not have some device on the inside of the network to wake it, then the next option is using a linux-based wireless access point or router on the inside to do the same.

 

Wake on LAN is actually very simple

 

https://en.wikipedia.org/wiki/Wake-on-LAN#Magic_packet

 

And because of that simplicity, you need only to know the MAC address of the device you want to wake.

 

Now caveat:

 

- Will not work on WiFi in most cases

- Will not work on Bluetooth

- Will not work on WWAN (eg 3G/4G/5G modems)

 

It's only guaranteed to work if the magic packet is sent from the same wired LAN that can see the MAC address, so if you have a bunch of unmanaged switches, that should be OK, but if you have managed switches or a router that partitions your network between wireless and wifi, or between your home and a public AP, or basically anything that prevents WiFi devices from seeing the wired connections, you can't send it from the wireless partition. 

 

Like if you do "arp -a" at the commandline you can have the MAC address's listed and if you don't see it there, then it's not found.

 

Now to open the other can of worms.

 

Assuming you've done things correctly, you only want to wake the system up via WoL. There is another thing called "WoL with PXE", which means a network payload that the DHCP server assigns is sent to the device to boot. This is completely viable to reinstall OS's with, which is why you want to make sure that PXE is turned off if the machine already has an OS installed to it. The likeliness that someone can just jack a machine and send a Remote access tool via PXE is not entirely impossible, but if someone has enough access to trigger Wake-on-LAN, then they certainly can also trigger PXE. Though it would probably have to be done by crashing the DHCP service on the router/modem.

 

Now if Wake on Lan is not the feature you want, there's also just setting the PC's BIOS to turn on the system at a set time, in which case it would be available over RDP (if it's windows) or VNC (if it's a Mac) or SSH (any/all three), however this still requires configuring your router so that you can reach it. so you might setup port 12322 to map to ssh port 22 on your first PC, and 12323 to map to the ssh port 22 on the second pc and so forth. Not the most secure option, but if SSH is available, you can also shut the machine down when you're done with it, without needing the desktop UI.

 

With that said, the wake-on-lan would let you do this at any point, just that since it operates at L2, it's not reachable from the public internet if the modem doesn't let it. So hence the rPi or something.

 

One final option, and this only really applies to servers and workstations, is if there is an BMC installed for IPMI support. Like Wake-on-Lan, you would want to connect to a device that is inside the network to do any IPMI access, because it's super-dangerous to expose IPMI to the internet, as that will not only give any potential hacker the ability to turn the machine on and off, but can also remotely administrate it via a BMC KVM feature.  Most desktops do not have this, but just something to keep in mind. The more expensive feature sets often do have these features for remote administration purposes.

 

[kpluck]

Don't a lot of BIOS settings include automatically powering up after a power failure? If the hardware in question has that feature, could you plug the system into a smartplug and power down/up the system that way? Killing the power would be interpreted as a power failure and turning it back on should then cause the system to automatically boot?

 

-kp