Reputation
Top tag
Next privilege 200 Rep.
See reduced ads
Badges
1 1 10
Impact
~23k people reached

  • 0 posts edited
  • 0 helpful flags
  • 16 votes cast
Nov
3
awarded  Famous Question
Aug
9
awarded  Popular Question
Jul
2
awarded  Curious
Feb
16
awarded  Popular Question
Feb
20
awarded  Notable Question
Oct
24
awarded  Popular Question
May
26
comment SLAAC and DNS, no really - what's the deal?
So why did they include SLAAC, then?
May
24
comment How reliable are IP address-based firewall restrictions?
Not just syn flood. Also reverse shells and the like.
May
24
comment SLAAC and DNS, no really - what's the deal?
The way I've read things is that DHCPv6 was a workaround introduced after the fact, and the ideal situation is to use "pure" IPv6 autoconfig.
May
24
comment Secure every connection between endpoints
Fair enough. Please see my comment about about the impacts of encryption upon HPC applications.
May
24
comment Secure every connection between endpoints
Wondering why the downvote? Did I misunderstand the question?
May
24
comment how to limit the number of simultaneous processes per user in the system?
Also /etc/login.conf in BSDs. I don't believe either of these are PAM only, but could be mistaken.
May
24
comment Secure every connection between endpoints
VPN is not meant for LAN traffic. Please see my response below. Also, what protocol is being used between the nodes? Since you mention "cluster", I'm going to assume HPC and advise you to be very cautious about encrypting inter-node comms as the latency can completely negate the advantages of parallel computing for many applications.
May
24
answered Secure every connection between endpoints
May
24
awarded  Commentator
May
24
comment How reliable are IP address-based firewall restrictions?
You can spoof an IP, you just can't easily get a response to your original packet. As I mentioned above, this might not be necessary to the exploit.
May
24
comment How reliable are IP address-based firewall restrictions?
Want to reinforce and add to this point. Something I don't see anyone else mentioning is that if the RDP server has a vulnerability, it's open for anyone on the internet to exploit. They might not need a two-way connection to do this, a simple payload inject/buffer overflow-type drive-by attack might be enough to force an outgoing connection from the RDP server back to the attacker. Okay, so this sounds a little far fetched, but the point is that you don't know what vulnerabilities might show up. Better to rely on methods of identification and authorization that are actually authenticated.
May
24
suggested rejected edit on How long should a password be, for filesystem encryption?
May
24
answered How long should a password be, for filesystem encryption?
May
24
accepted Protect files on NTFS volume from Domain Administrators