Would like some critical analysis of my 1st pix config

[samurah]

hi guys hope the experts in pix will give good advice on my very first pix config:

scenario: 3 servers, 2 win2k's and 1 linux
1st win2k running web server
2nd win2k server running sql server
linux running web server and mail server

a nat'ed network has been setup

here is my config

Building configuration...
: Saved
:
PIX Version 6.2(2)
nameif ethernet0 outside security0
nameif ethernet1 inside security100
enable password xxxxxxxxxxxxxx encrypted
passwd xxxxxxxxxxxxxx encrypted
hostname pix
domain-name mydom.com
fixup protocol ftp 21
fixup protocol http 80
fixup protocol h323 h225 1720
fixup protocol h323 ras 1718-1719
fixup protocol ils 389
fixup protocol rsh 514
fixup protocol rtsp 554
fixup protocol smtp 25
fixup protocol sqlnet 1521
fixup protocol sip 5060
fixup protocol skinny 2000
fixup protocol http 8443
fixup protocol http 8070
names
name 192.168.73.3 win2k
name 192.168.73.4 linux
name 192.168.73.5 sql
pager lines 24
logging on
interface ethernet0 auto
interface ethernet1 auto
mtu outside 1500
mtu inside 1500
ip address outside 192.168.73.2 255.255.255.0
ip address inside 172.25.1.1 255.255.255.255
ip audit info action alarm
ip audit attack action alarm
pdm location 172.25.1.3 255.255.255.255 inside
pdm location 172.25.1.4 255.255.255.255 inside
pdm location 172.25.1.5 255.255.255.255 inside
pdm logging emergencies 100
pdm history enable
arp timeout 14400
global (outside) 1 win2k-192.168.73.255
nat (inside) 1 0.0.0.0 0.0.0.0 0 0
static (inside,outside) win2k 172.25.1.3 netmask 255.255.255.255 0 0
static (inside,outside) linux 172.25.1.4 netmask 255.255.255.255 0 0
static (inside,outside) sql 172.25.1.5 netmask 255.255.255.255 0 0
conduit permit tcp host 172.25.1.3 eq

http://www any

conduit permit tcp host 172.25.1.3 eq ftp any
conduit permit tcp host 172.25.1.4 eq

http://www any

conduit permit tcp host 172.25.1.4 eq ftp any
conduit permit tcp host 172.25.1.4 eq smtp any
conduit permit tcp host 172.25.1.4 eq pop3 any
route outside 0.0.0.0 0.0.0.0 192.168.73.1 1
timeout xlate 3:00:00
timeout conn 1:00:00 half-closed 0:10:00 udp 0:02:00 rpc 0:10:00 h323 0:05:00 sip 0:30:00

sip_media 0:02:00
timeout uauth 0:05:00 absolute
aaa-server TACACS+ protocol tacacs+
aaa-server RADIUS protocol radius
aaa-server LOCAL protocol local
http server enable
http 172.25.1.0 255.255.255.0 inside
no snmp-server location
no snmp-server contact
snmp-server community public
no snmp-server enable traps
floodguard enable
no sysopt route dnat
ssh 172.25.1.0 255.255.255.0 inside
ssh timeout 5
username admin password xxxxxxxxxxxxxx encrypted privilege 2
terminal width 80
Cryptochecksum:2352f120c6cda4f71bc7f6fec22218fe
: end
[OK]

 

[tbissett]

A few things:

1) Your global statement will not work as is. You cannot assigned a static and global to the same IP (as you did with Win2K, linux, and sql). Also, 192.168.73.255 is a broadcast address, so it can't be assigned as part of a global either. Instead of doing an entire range of IPs, just use the outside interface IP for your global segment with the following command:
global (outside) 1 interface
Don't forget to add a reverse-dns entry on the outside for this IP. That helps with FTP and HTTP performance.

2) I would recommend against using conduits as Cisco has stated they will be phasing them out eventually. Instead, use the access-list command. Here's how to do the same thing with the access-lists:
access-list inbound permit tcp any host win2k eq www
access-list inbound permit tcp any host win2k eq ftp
access-list inbound permit tcp any host linux eq www
access-list inbound permit tcp any host linux eq ftp
access-list inbound permit tcp any host linux eq smtp
access-list inbound permit tcp any host linux eq pop3
access-group inbound in interface outside

3) Instead of NAT (inside) 1 0.0.0.0 0.0.0.0, Specify the actual IPs to be NATed. This helps in preventing IP spoofing attempts going OUT through your firewall. Why is this important? Because users tend to download anything and everything, and it's not uncommon for trojans, DoS agents, etc. to be installed on your network. Those agents usually spoof their source IP. With a NAT (inside) 1 0 0 command, those packets would be allowed out, thereby allowing your PCs to participate in these attacks. Instead, change the NAT command to something like:
nat (inside) 1 172.25.1.0 255.255.255.0
You can add additional networks just by using the same process ID (the number 1 above).

4) Use an outbound access-list to deny all traffic unless you specifically permit it. Same reasons as Number 3, but it also is a good way for you to keep track of exactly what is going through your firewall. Something like:
access-list outbound permit tcp any any eq www
access-list outbound permit tcp any any eq ftp
access-list outbound permit tcp any any eq telnet
access-list outbound permit udp any any eq domain
access-group outbound in interface inside

5) Turn off pings to the outside interface. Just a security best practice:
icmp deny any outside

6) Since there is no access from the outside directly to the SQL server, you might consider moving that box to the INSIDE segment and adding the rules to permit your DMZ host to talk to it. Right now, if someone broke into any machine on your DMZ, they could use it to go after your SQL server on any protocol and any port. If you moved the SQL box to the inside, the hacker would first have to break into the specific machine on the DMZ that is allowed to talk to the SQL server. Then, the intruder would have only one or two ports in which to try to break into the SQL server (not to mention going back through the firewall and it's stateful inspection one more time).

Hope this helps...