Re: Remote terminal service - Comments
- From: "Frankster" <Frank@xxxxxxxxxxxxxx>
- Date: Tue, 20 Dec 2005 19:13:28 -0700
Thanks for your answer Jeff. I'm still having a little difficulty
understanding just what you might recommend. "Poking a hole" ,as you call
it, is not what I call it when you do port forwarding from your firewall to
a single box behind your firewall, that is NAT'd.
Anyhow, near as I can tell, you're thinking VPN or DMZ, pretty much period.
Is that right?
Incidentally, about DMZ. I believe in DMZs wholeheartedly, if your network
warrants the expense. But if you "poke holes" from your DMZ to allow
connectivity to your internal network, what's the point in having that other
router for your DMZ anyway? Many folks do DMZs this way, for better or
worse.
I view "publicly accessible" servers different from "directly on the
Internet". All the web browsing everyone does is to "public accessible"
servers. But not all are "directly on the Internet".
-Frank
"Jeff Pitsch" <jeff@xxxxxxxxxxxxxxxxx> wrote in message
news:ubiW6JaBGHA.2668@xxxxxxxxxxxxxxxxxxxxxxx
> Obviously this is completely my opinion so take it for what it's worth.
>
> An internal server should never be exposed directly to the internet. This
> is regardless of a NAT because it still allows direct access to the
> server. The preferable solution is something like a VPN (a traditional VPN
> does typically slow things down but it shouldn't be that bad) or a man in
> the middle (citrix Secure Gateway). It comes down to making it that much
> harder for someone to get into your internal network. for example, a DOS
> attack. If you do strictly the NAT, your still exposed your internal
> network and if an attack isn't shut down, you've now compromised your
> internal network. if you do a man in the middle (2x loadbalancer or
> Citrix Secure Gateway), the attack is directed at that one box in the DMZ.
> You've effectively cut off the initial attack.
>
> Now on top of that, if you are going to poke a hole to your internal
> network, why have a firewall at all? What good does it do if you're
> exposing your internal network? Yes someone could hack the DMZ server and
> then hack into your internal network but, again, your putting another
> layer of defense in. This lends itself to you may be able to detect an
> intruder in your DMZ before they get inside.
>
> to me this is just the #1 tenant of security, you never, ever directly
> expose your internal network.
>
> as well, not everyone is going to hacked and not everyone is even going to
> have an attempt. That's great and that's the way it should be. On the
> same token (to take away from mothers everywhere), if everyone jumped off
> a bridge would you? to paraphrase Jurassic Park, you were so busy
> figuring out if you could, you never thought about if you should.
>
> does any of that make sense?
>
> Jeff Pitsch
> http://www.sbcgatekeeper.com
> Your Terminal Services Security Website
>
> "Frankster" <Frank@xxxxxxxxxxxxxx> wrote in message
> news:q-KdnZ6l1r9tsjXeRVn-qQ@xxxxxxxxxxxxxxx
>>> the site has only been up for a few months. It takes time for content.
>>
>> Yeah, I know :-)
>>
>> One thing I was looking for was your recommended method of not placing a
>> TS (or any server) "directly" on the Internet.
>>
>> I've been running a TS "directly" on the Internet for years without
>> issue. First NT, then 2000, now 2003. But, my definition of "directly" is
>> probably different than yours. Mine is behind a network firewall, using a
>> private IP after being NAT'd. Still, no VPN or other security measures
>> (except NATing) have been taken. Yes, a complex password requirement for
>> all users on the box is a given. Perhaps this qualifies in your
>> definition of "not directly". I dunno...
>>
>> OTOH, I have also often ran a TS within a VPN, NAT'd and with RSA
>> SecureID. Works too. But slower. Perhaps this is your definition of "not
>> directly".
>>
>> I would take OS hardening to be a given in all public access situations.
>>
>> Can you comment?
>>
>> -Frank
>>
>
>
.
- Follow-Ups:
- Re: Remote terminal service - Comments
- From: Jeff Pitsch
- Re: Remote terminal service - Comments
- References:
- Remote terminal service - Comments
- From: myname
- Re: Remote terminal service - Comments
- From: Jeff Pitsch
- Re: Remote terminal service - Comments
- From: Frankster
- Re: Remote terminal service - Comments
- From: Jeff Pitsch
- Re: Remote terminal service - Comments
- From: Frankster
- Re: Remote terminal service - Comments
- From: Jeff Pitsch
- Re: Remote terminal service - Comments
- From: Frankster
- Re: Remote terminal service - Comments
- From: Jeff Pitsch
- Remote terminal service - Comments
- Prev by Date: Re: How hard is it to config Win2003 to serve a single app and offer users little else?
- Next by Date: Re: REMOTE CONTROL (W2K3) stops working after updating WXP Pro SP1 to
- Previous by thread: Re: Remote terminal service - Comments
- Next by thread: Re: Remote terminal service - Comments
- Index(es):
Relevant Pages
|