This is a three phased approach:
1. How can you protect data by establishing security on NFS mounts.
2. How can you possibly work around in hacking that data still - a security breach so to speak.
3. Knowing a security breach occured what methods can be employed to further tighten it.
So on highlevel -
NFS shared mount access can be restricted to IP address's.
Using NMAP Scans - we can obtain information on the neighbouring hosts on the same subnets.
By spinning up Virtual IP's and arpspoofs/arpings and also using NAT rules the server can still be accessed.
The level of granularity in securing this attack can be blocked by controling/using MAC address association - this can be accomplished using EBTABLES.
I have given a detailed picture of how NFS hacking can be done. i.e Step 1,2,3 described above already on my previous article -'NFS Hacking using NAT (Network-Address-Translation)' I shall still capture the steps here, but focus or goal is to how to accomplish #3.
So the basic setup remains same. RBNETOPS will the NFS Server which is used to spin up the shared mounts RBNETAPP and RBNETLAB are two clients given access to mount their respective shares with access restricted by it's own IP Addressess. RBNETLAB will then be considered as a malicious host - who attempts to do network scan to identify the hosts around - finds and attempts to mount the directory owned by RBNETAPP and successfully mounts it.
System admins at RBNETOPS later realizes that there's a security breach and a gap that need to filled up.This is where EBTABLES will come into play.
We need bridging because ultimately we need to setup ebtable rules which solely works on ETHERNET bridging. Bridge as the name implies acts as a bridge between two ethernets for packet traversal. Usually bridging is done by physical devices but can also be setup within the Linux kernel using bridge modules. So to start with we need to install the bridge utils RPM.
By default bridges are not configured and ebtables can only work on ETHERNET bridges. So first - add the bridge. At this point a bridge is created by no interfaces/IP addresses are tied to it.
yum install bridge-utils
[root@rbnetops ~]# brctl addbr brnfs
[root@rbnetops ~]# brctl show
bridge name bridge id STP enabled interfaces
brnfs 8000.000000000000 no
[root@rbnetops network-scripts]# ifconfig
enp0s3: flags=4163 mtu 1500
inet 192.168.56.151 netmask 255.255.255.0 broadcast 192.168.56.255
inet6 fe80::a00:27ff:fe98:3854 prefixlen 64 scopeid 0x20
ether 08:00:27:98:38:54 txqueuelen 1000 (Ethernet)
RX packets 658 bytes 89567 (87.4 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 221 bytes 35688 (34.8 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp0s8: flags=4163 mtu 1500
inet 192.168.2.151 netmask 255.255.255.0 broadcast 192.168.2.255
inet6 fe80::a00:27ff:fecb:b503 prefixlen 64 scopeid 0x20
ether 08:00:27:cb:b5:03 txqueuelen 1000 (Ethernet)
RX packets 14904 bytes 1963214 (1.8 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1088 bytes 95241 (93.0 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
...
output trimmed to omit loopback interface..
Ideally bridging is done using two or more physical devices. In my case 192.168.56.151 is an interface defined as a 'Host-Only-Adapter' within VirtualBox and 192.168.2.151 is a 'Bridged Adapter' - basically I bridged that to my physical interface. Since these aren't same type of NIC's I wanted to avoid mixing up/using both NIC's to create the bridge. Hence going with only one enp0s8 here for the purpose of demo. The IP address of the NIC has to be set to 0.0.0.0
[root@rbnetops network-scripts]# ifconfig enp0s8 0.0.0.0 down
[root@rbnetops network-scripts]# brctl addif brnfs enp0s8
[root@rbnetops network-scripts]# brctl show
bridge name bridge id STP enabled interfaces
brnfs 8000.080027cbb503 no enp0s8
[root@rbnetops network-scripts]# ifconfig enp0s8 up
[root@rbnetops network-scripts]# ifconfig
enp0s3: flags=4163 mtu 1500
inet 192.168.56.151 netmask 255.255.255.0 broadcast 192.168.56.255
inet6 fe80::a00:27ff:fe98:3854 prefixlen 64 scopeid 0x20
ether 08:00:27:98:38:54 txqueuelen 1000 (Ethernet)
RX packets 980 bytes 122441 (119.5 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 464 bytes 70011 (68.3 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp0s8: flags=4163 mtu 1500
inet6 fe80::a00:27ff:fecb:b503 prefixlen 64 scopeid 0x20
ether 08:00:27:cb:b5:03 txqueuelen 1000 (Ethernet)
RX packets 16168 bytes 2076355 (1.9 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1098 bytes 96074 (93.8 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
...
output trimmed...
Bring the bridge interface up, you can see it's up and RUNNING although there isn't an IP address associated with it still. brctl command has options which can be used to show what MAC addresses are being discovered by the bridge that it can communicate with.
[root@rbnetops network-scripts]# ifconfig brnfs up
[root@rbnetops network-scripts]# ifconfig brnfs
brnfs: flags=4163 mtu 1500
inet6 fe80::a00:27ff:fecb:b503 prefixlen 64 scopeid 0x20
ether 08:00:27:cb:b5:03 txqueuelen 0 (Ethernet)
RX packets 337 bytes 51239 (50.0 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 8 bytes 648 (648.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
[root@rbnetops network-scripts]# brctl showmacs brnfs
port no mac addr is local? ageing timer
1 00:1f:bc:02:6f:11 no 135.92
1 08:00:27:cb:b5:03 yes 0.00
1 08:00:27:cb:b5:03 yes 0.00
1 5c:d9:98:60:6a:1e no 0.32
Now attach an IP address to the bridge interface.
[root@rbnetops network-scripts]# ifconfig brnfs 192.168.2.151
[root@rbnetops network-scripts]# ifconfig
brnfs: flags=4163 mtu 1500
inet 192.168.2.151 netmask 255.255.255.0 broadcast 192.168.2.255
inet6 fe80::a00:27ff:fecb:b503 prefixlen 64 scopeid 0x20
ether 08:00:27:cb:b5:03 txqueuelen 0 (Ethernet)
RX packets 768 bytes 72295 (70.6 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 17 bytes 2837 (2.7 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp0s3: flags=4163 mtu 1500
inet 192.168.56.151 netmask 255.255.255.0 broadcast 192.168.56.255
inet6 fe80::a00:27ff:fe98:3854 prefixlen 64 scopeid 0x20
ether 08:00:27:98:38:54 txqueuelen 1000 (Ethernet)
RX packets 1123 bytes 138298 (135.0 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 576 bytes 86993 (84.9 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
enp0s8: flags=4163 mtu 1500
inet6 fe80::a00:27ff:fecb:b503 prefixlen 64 scopeid 0x20
ether 08:00:27:cb:b5:03 txqueuelen 1000 (Ethernet)
RX packets 16941 bytes 2160167 (2.0 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 1117 bytes 99063 (96.7 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
Now move onto RBNETLAB and configure the VIP (Virutal IP ) - using the IP address of RBNETAPP's machine. If you have logged in using putty/ssh to RBNETAPP using the ip address 192.168.2.152 you will potentially loose connectivity because of IP address conflicts now on the same subnet.
[root@rbnetlab ~]# ifconfig enp0s8:1 192.168.2.152 netmask 255.255.255.0 up
[root@rbnetlab ~]# ifconfig enp0s8
enp0s8: flags=4163 mtu 1500
inet 192.168.2.153 netmask 255.255.255.0 broadcast 192.168.2.255
inet6 fe80::a00:27ff:feb3:8bb2 prefixlen 64 scopeid 0x20
ether 08:00:27:b3:8b:b2 txqueuelen 1000 (Ethernet)
RX packets 16507 bytes 1074253 (1.0 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 740 bytes 97038 (94.7 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
[root@rbnetlab ~]# ifconfig enp0s8:1
enp0s8:1: flags=4163 mtu 1500
inet 192.168.2.152 netmask 255.255.255.0 broadcast 192.168.2.255
ether 08:00:27:b3:8b:b2 txqueuelen 1000 (Ethernet)
[root@rbnetlab ~]# sysctl net.ipv4.ip_nonlocal_bind=1
net.ipv4.ip_nonlocal_bind = 1
[root@rbnetlab ~]# echo "1" > /proc/sys/net/ipv4/ip_forward
[root@rbnetlab ~]# arpspoof -i enp0s8 -t 192.168.2.151 192.168.2.152
Here are the steps to do NAT'ing otherwise we need to mangle the packets.
There are two steps here. One is doing Source NAT and another is Destination NAT. Source NAT is done in POSTROUTING chain so that the packets right before packets were being sent out by kernel it's mangled to make sure the outgoing IP address is 192.168.2.152 ( Virutal IP).
Same is the case with the PREROUTING chain, any incoming packets received are intercepted and mangled so that the local kernel processes once they receive them will see it as a packet came from 192.168.2.152.
Starting with Redhat/CentOS 7.x - the legacy IPTABLES are being taken over by firewalld. Basically firewalld offers lot of efficiency and it's wrapped around the legacy iptables so still allows to use the iptable rules to be defined. That's done through 'passthrough' functionality.
I ran into some issues with PUBLIC zone of firewalld, so switched to 'trusted' zone temporarily. Yet to figure out the reason as to why it didn't work but for now going with 'trusted' zone.
firewall-cmd --direct --passthrough ipv4 -t nat -I POST_trusted -s 192.168.2.153 -j SNAT --to 192.168.2.152 firewall-cmd --direct --passthrough ipv4 -t nat -I PRE_trusted -s 192.168.2.152 -j DNAT --to 192.168.2.153 [root@rbnetlab ~]# firewall-cmd --direct --passthrough ipv4 -t nat -I POST_trusted -s 192.168.2.153 -j SNAT --to 192.168.2.152 success [root@rbnetlab ~]# firewall-cmd --direct --passthrough ipv4 -t nat -I PRE_trusted -s 192.168.2.152 -j DNAT --to 192.168.2.153 success [root@rbnetlab ~]# firewall-cmd --zone=trusted --change-interface=enp0s8 success [root@rbnetlab ~]# firewall-cmd --get-zone-of-interface=enp0s8 trusted [root@rbnetlab ~]# firewall-cmd --zone=trusted success [root@rbnetlab ~]# firewall-cmd --get-default-zone trusted
Here is a snap of iptables rules once it's been added.
[root@rbnetlab ~]# iptables -t nat -L Chain PREROUTING (policy ACCEPT) target prot opt source destination PREROUTING_direct all -- anywhere anywhere PREROUTING_ZONES_SOURCE all -- anywhere anywhere PREROUTING_ZONES all -- anywhere anywhere Chain INPUT (policy ACCEPT) target prot opt source destination Chain OUTPUT (policy ACCEPT) target prot opt source destination OUTPUT_direct all -- anywhere anywhere Chain POSTROUTING (policy ACCEPT) target prot opt source destination POSTROUTING_direct all -- anywhere anywhere POSTROUTING_ZONES_SOURCE all -- anywhere anywhere POSTROUTING_ZONES all -- anywhere anywhere Chain OUTPUT_direct (1 references) target prot opt source destination Chain POSTROUTING_ZONES (1 references) target prot opt source destination POST_trusted all -- anywhere anywhere POST_trusted all -- anywhere anywhere POST_trusted all -- anywhere anywhere Chain POSTROUTING_ZONES_SOURCE (1 references) target prot opt source destination Chain POSTROUTING_direct (1 references) target prot opt source destination Chain POST_trusted (3 references) target prot opt source destination SNAT all -- rbnetlab anywhere to:192.168.2.152 POST_trusted_log all -- anywhere anywhere POST_trusted_deny all -- anywhere anywhere POST_trusted_allow all -- anywhere anywhere Chain POST_trusted_allow (1 references) target prot opt source destination Chain POST_trusted_deny (1 references) target prot opt source destination Chain POST_trusted_log (1 references) target prot opt source destination Chain PREROUTING_ZONES (1 references) target prot opt source destination PRE_trusted all -- anywhere anywhere PRE_trusted all -- anywhere anywhere PRE_trusted all -- anywhere anywhere Chain PREROUTING_ZONES_SOURCE (1 references) target prot opt source destination Chain PREROUTING_direct (1 references) target prot opt source destination Chain PRE_trusted (3 references) target prot opt source destination DNAT all -- rbnetapp anywhere to:192.168.2.153 PRE_trusted_log all -- anywhere anywhere PRE_trusted_deny all -- anywhere anywhere PRE_trusted_allow all -- anywhere anywhere Chain PRE_trusted_allow (1 references) target prot opt source destination Chain PRE_trusted_deny (1 references) target prot opt source destination Chain PRE_trusted_log (1 references) target prot opt source destination
Let's attempt the hack again basically to see if the mounts are working fine.
[root@rbnetlab ~]# mount -t nfs -o addr=192.168.2.152 rbnetops:/u01/data/nfsshare_rbnetapp /u01/hackit/ -v mount.nfs: timeout set for Wed May 6 14:44:08 2015 mount.nfs: trying text-based options 'vers=4,addr=192.168.2.151,clientaddr=192.168.2.153' [root@rbnetlab ~]# cd /u01/hackit/ [root@rbnetlab hackit]# ls -ltr total 4 -rw-r--r--. 1 root root 32 Apr 2 13:24 rbnetapp.dat [root@rbnetlab hackit]# cat rbnetapp.dat Confidential created @ rbnetapp
Configure firewalld on RBNEOPS. This is required for setting up EBTABLES. May be there's a way to setup the EBTABLE rules without the passthrough option (yet to explore), but with quite a few hurdles/learning this is the only way I know now! The main thing to note here is IP/MAC association in the rule. So basically when a request is triggered with our NAT and arpspoofing from RBNETLAP while it hits RBNETOPS the ebtable rules kicks in and rules checks this association and will take the action according to the last parameter ( -j ACCEPT or -j DROP).
systemctl start firewalld
systemctl status firewalld
systemctl start ebtables
[root@rbnetops network-scripts]# systemctl start firewalld
[root@rbnetops network-scripts]# systemctl status firewalld
firewalld.service - firewalld - dynamic firewall daemon
Loaded: loaded (/usr/lib/systemd/system/firewalld.service; enabled)
Active: active (running) since Wed 2015-05-06 14:51:08 EDT; 14ms ago
Main PID: 14127 (firewalld)
CGroup: /system.slice/firewalld.service
ââ14127 /usr/bin/python -Es /usr/sbin/firewalld --nofork --nopid
ââ14181 /usr/bin/python -Es /usr/sbin/firewalld --nofork --nopid
May 06 14:51:08 rbnetops.intercloudzone.com systemd[1]: Starting firewalld - dynamic firewall d.....
May 06 14:51:08 rbnetops.intercloudzone.com systemd[1]: Started firewalld - dynamic firewall daemon.
Hint: Some lines were ellipsized, use -l to show in full.
[root@rbnetops network-scripts]# ebtables -L --Lc
Bridge table: filter
Bridge chain: INPUT, entries: 0, policy: ACCEPT
Bridge chain: FORWARD, entries: 0, policy: ACCEPT
Bridge chain: OUTPUT, entries: 0, policy: ACCEPT
[root@rbnetops network-scripts]# systemctl start ebtables
On RBNETOPS run the below..
firewall-cmd --direct --passthrough eb -A INPUT -p IPv4 -s 08:00:27:85:30:eb --ip-src 192.168.2.152 -j ACCEPT
firewall-cmd --direct --passthrough eb -A INPUT -p IPv4 -s 08:00:27:b3:8b:b2 --ip-src 192.168.2.153 -j ACCEPT
firewall-cmd --direct --passthrough eb -A INPUT -p IPv4 --limit 5/min --limit-burst 2 --log-level notice --log-prefix "EBTABLE RULE VIOLATION" --log-ip --log-arp -j DROP
firewall-cmd --direct --passthrough eb -A INPUT -p IPv4 -j DROP
[root@rbnetops network-scripts]# ebtables -L --Lc
Bridge table: filter
Bridge chain: INPUT, entries: 4, policy: ACCEPT
-p IPv4 -s 8:0:27:85:30:eb --ip-src 192.168.2.152 -j ACCEPT , pcnt = 0 -- bcnt = 0
-p IPv4 -s 8:0:27:b3:8b:b2 --ip-src 192.168.2.153 -j ACCEPT , pcnt = 0 -- bcnt = 0
-p IPv4 --limit 5/min --limit-burst 2 --log-level notice --log-prefix "EBTABLE RULE VIOLATION" --log-ip --log-arp -j DROP , pcnt = 0 -- bcnt = 0
-p IPv4 -j DROP , pcnt = 0 -- bcnt = 0
Bridge chain: FORWARD, entries: 0, policy: ACCEPT
Bridge chain: OUTPUT, entries: 0, policy: ACCEPT
Now that rules are in place, attempt a mount again. The mount get's hung / stuck at the prompt...
[root@rbnetlab ~]# mount -t nfs -o addr=192.168.2.152 rbnetops:/u01/data/nfsshare_rbnetapp /u01/hackit/ -v mount.nfs: timeout set for Wed May 6 14:57:49 2015 mount.nfs: trying text-based options 'vers=4,addr=192.168.2.151,clientaddr=192.168.2.153'
Now to look at or to find why the mount isn't happening, log back into RBNETOPS and see what's happening. You can see the packets are dropped by ebtables and if you refer back to the rules there's a log prefix added " --log-prefix "EBTABLE RULE VIOLATION" which is what you can see in the logs as comments for every packet being dropped. The limit-burst and limits on the rules prevent to flood the logs with repeated lines.
[root@rbnetops network-scripts]# tail -f /var/log/messages May 6 14:54:57 rbnetops kernel: nf_conntrack version 0.5.0 (16384 buckets, 65536 max) May 6 14:54:57 rbnetops ebtables: [ OK ] May 6 14:54:57 rbnetops kernel: Ebtables v2.0 unregistered May 6 14:54:57 rbnetops systemd: Stopped Ethernet Bridge Filtering tables. May 6 14:54:57 rbnetops kernel: ip6_tables: (C) 2000-2006 Netfilter Core Team May 6 14:54:57 rbnetops kernel: Ebtables v2.0 registered May 6 14:54:57 rbnetops systemd: Started firewalld - dynamic firewall daemon. May 6 14:55:01 rbnetops systemd: Created slice user-992.slice. May 6 14:55:01 rbnetops systemd: Starting Session 59 of user pcp. May 6 14:55:01 rbnetops systemd: Started Session 59 of user pcp. May 6 14:55:44 rbnetops chronyd[685]: Can't synchronise: no reachable sources May 6 14:55:49 rbnetops kernel: EBTABLE RULE VIOLATION IN=enp0s8 OUT= MAC source = 08:00:27:b3:8b:b2 MAC dest = 08:00:27:cb:b5:03 proto = 0x0800 IP SRC=192.168.2.152 IP DST=192.168.2.151, IP tos=0x00, IP proto=6 SPT=999 DPT=2049 May 6 14:55:50 rbnetops kernel: EBTABLE RULE VIOLATION IN=enp0s8 OUT= MAC source = 08:00:27:b3:8b:b2 MAC dest = 08:00:27:cb:b5:03 proto = 0x0800 IP SRC=192.168.2.152 IP DST=192.168.2.151, IP tos=0x00, IP proto=6 SPT=999 DPT=2049