There is a recent case whereby it was reported that one of the production servers was hacked on port 1521. However, I am not sure how this was possible, as I checked that the OS firewall (iptables) is on :
Port 1521 not open :
However, it is listening on port 1521 :
I assume it is listening because the application is turned on, and thus the service related to port 1521 turned on and that is why it is listening on this port?
Even if the network firewall (physical) is open, if the iptables is running in the server, it should not allow port 1521 to be open/listening?
I can't think of any other way how people can hack the server on port 1521. Please help clear my doubt.
If i remember correctly 1521 is one of the standard ports for the Oracle listener, so i suppose you have an Oracle database running there. That the listener listens is quite as it should be, no?
What makes you think the server "was hacked"?
I mean, iptables is just a packet filter and as such it cannot discern between legitimate content and an illegitimate one. It filters packets based on IP address (layer 3) and port (layer 4), nothing more, nothing less. Obviously you need to allow traffic to the configured port of the listener otherwise the database would not be usable. So either you allow this port or you disable it (eventually restricting to a certain range of IP addresses), but what content goes over this port (i.e. legitimate database queries vs. malicious content) the packet filter is the wrong tool to assess. For that you will need a "stateful inspection" type of firewall which iptables is not.
Also be aware that the concept of "host based firewalls" is a flawed one per design. A hosts role is either providing a service (that is: some application) OR providing firewall services, but not both! The reason is you don't want the host you want to protect run the firewall itself, beause in this scenario the malicious packages already have reached the interface they are trying to attack. You want the firewall in front of (and separated from) the host you try to protect so that the malicious content doesn't even reach the interface you want to protect.
I hope this helps.
bakunin
Last edited by bakunin; 02-04-2019 at 05:08 PM..
Reason: confused "stateful inspection" with "deep state inspection" - oh, the paranoia
Hi Guys,
My certificate in /etc/ssh is different to what is on port 22.
username@server:~$ ssh-keyscan -p 22 127.0.0.1 > /tmp/rsa.tmp
# 127.0.0.1 SSH-1.99-OpenSSH_33.33
username@server:~$ ssh-keygen -lf /tmp/rsa.tmp
1024 46:something..................... 127.0.0.1... (0 Replies)
Hello!
First of all: I am a newbie. :o :(
I have a CentOS 64bit server with Plesk Panel 8.6.
And have been hacked. :mad:
After many tries and support tickets, I am configuring a new server, with Suse 11 and Plesk 9.2.
I know that Plesk 8.6 have a backup utility (Parallels Plesk Control... (3 Replies)
One of the most important ways to keep tou machine secure is to know when it has been broken into. The less time hackers have on your system, the less they can do to it, and the greater you chancens of kicking them off and repairing the damage.
The more sophisticated the hacker, the less likely... (8 Replies)
Hello there,
I have a big problem, and I hope somebody can help me. I try to realize a port forward over three server. Here is a picture...
Client Server1 | Server2
------- ------- | -------
|...... | |...... | | |...... ... (2 Replies)
Just got old ATT server (10 base T)shipped and want to connect to Windows using com port. Got hardware to connect RJ45 from windows box & serial on ATT. I added XP static ip to host file but get no ping return. Do I have to open unix com port? How? (2 Replies)
Hi,
i think someone has hacked my server, the following rules used to come which i haven't put. Please help me i couldnt find out how this rules are apply,
i think someone has put an script which generates enables the rules.
But after restarting the iptables everything seems to be working... (0 Replies)
I want to write a server application that would accept HTTP requests from client.
The server would be on a machine that has no connection to the INTERNET.
The clients that would be posting their HTTP requests would be doing so through webbrowser .Thus it would be sort of intranet application.... (0 Replies)