4 include::attributes.txt[]
9 pve-firewall - The PVE Firewall Daemon
15 include::pve-firewall.8-synopsis.adoc[]
25 include::attributes.txt[]
28 // Copied from pve wiki: Revision as of 08:45, 9 November 2015
30 Proxmox VE Firewall provides an easy way to protect your IT
31 infrastructure. You can easily setup firewall rules for all hosts
32 inside a cluster, or define rules for virtual machines and
33 containers. Features like firewall macros, security groups, IP sets
34 and aliases help making that task easier.
36 While all configuration is stored on the cluster file system, the
37 iptables based firewall runs on each cluster node, and thus provides
38 full isolation between virtual machines. The distributed nature of
39 this system also provides much higher bandwidth than a central
42 NOTE: If you enable the firewall, all traffic is blocked by default,
43 except WebGUI(8006) and ssh(22) from your local network.
49 The Proxmox VE firewall groups the network into the following logical zones:
53 Traffic from/to a cluster node
57 Traffic from/to a specific VM
59 For each zone, you can define firewall rules for incoming and/or
63 Ports used by Proxmox VE
64 ------------------------
67 * VNC Web console: 5900-5999
69 * sshd (used for cluster actions): 22
71 * corosync multicast (if you run a cluster): 5404, 5405 UDP
77 All firewall related configuration is stored on the proxmox cluster
78 file system. So those files are automatically distributed to all
79 cluster nodes, and the 'pve-firewall' service updates the underlying
80 iptables rules automatically on any change. Any configuration can be
81 done using the GUI (i.e. Datacenter -> Firewall -> Options tab (tabs
82 at the bottom of the page), or on a Node -> Firewall), so the
83 following configuration file snippets are just for completeness.
85 Cluster wide configuration is stored at:
87 /etc/pve/firewall/cluster.fw
89 The firewall is completely disabled by default, so you need to set the
94 # enable firewall (cluster wide setting, default is disabled)
98 The cluster wide configuration can contain the following data:
102 * Security group definitions
103 * Cluster wide firewall rules for all nodes
105 VM firewall configuration is read from:
107 /etc/pve/firewall/<VMID>.fw
109 and contains the following data:
113 * Firewall rules for this VM
114 * VM specific options
116 And finally, any host related configuration is read from:
118 /etc/pve/nodes/<nodename>/host.fw
120 This is useful if you want to overwrite rules from 'cluster.fw'
121 config. You can also increase log verbosity, and set netfilter related
124 Enabling the Firewall for VMs and Containers
125 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
127 You need to enable the firewall on the virtual network interface configuration
128 in addition to the general 'Enable Firewall' option in the 'Options' tab.
133 Firewall rules consists of a direction (`IN` or `OUT`) and an
134 action (`ACCEPT`, `DENY`, `REJECT`). You can also specify a macro
135 name. Macros contain predifined sets of rules and options. Rules can be disabled by prefixing them with '|'.
137 .Firewall rules syntax
141 DIRECTION ACTION [OPTIONS]
142 |DIRECTION ACTION [OPTIONS] # disabled rule
144 DIRECTION MACRO(ACTION) [OPTIONS] # use predefined macro
147 The following options can be used to refine rule matches.
149 include::pve-firewall-rules-opts.adoc[]
151 Here are some examples:
155 IN SSH(ACCEPT) -i net0
156 IN SSH(ACCEPT) -i net0 # a comment
157 IN SSH(ACCEPT) -i net0 -source 192.168.2.192 # only allow SSH from 192.168.2.192
158 IN SSH(ACCEPT) -i net0 -source 10.0.0.1-10.0.0.10 # accept SSH for ip range
159 IN SSH(ACCEPT) -i net0 -source 10.0.0.1,10.0.0.2,10.0.0.3 #accept ssh for ip list
160 IN SSH(ACCEPT) -i net0 -source +mynetgroup # accept ssh for ipset mynetgroup
161 IN SSH(ACCEPT) -i net0 -source myserveralias #accept ssh for alias myserveralias
163 |IN SSH(ACCEPT) -i net0 # disabled rule
165 IN DROP # drop all incoming packages
166 OUT ACCEPT # accept all outgoing packages
172 A security group is a collection of rules, defined at cluster level, which
173 can be used in all VMs' rules. For example you can define a group named
174 `webserver` with rules to open the http and https ports.
177 # /etc/pve/firewall/cluster.fw
180 IN ACCEPT -p tcp -dport 80
181 IN ACCEPT -p tcp -dport 443
184 Then, you can add this group to a VM's firewall
187 # /etc/pve/firewall/<VMID>.fw
197 IP Aliases allow you to associate IP addresses of networks with a
198 name. You can then refer to those names:
200 * inside IP set definitions
201 * in `source` and `dest` properties of firewall rules
203 Standard IP alias `local_network`
204 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
206 This alias is automatically defined. Please use the following command
207 to see assigned values:
210 # pve-firewall localnet
211 local hostname: example
212 local IP address: 192.168.2.100
213 network auto detect: 192.168.0.0/20
214 using detected local_network: 192.168.0.0/20
217 The firewall automatically sets up rules to allow everything needed
218 for cluster communication (corosync, API, SSH) using this alias.
220 The user can overwrite these values in the cluster.fw alias
221 section. If you use a single host on a public network, it is better to
222 explicitly assign the local IP address
225 # /etc/pve/firewall/cluster.fw
227 local_network 1.2.3.4 # use the single ip address
233 IP sets can be used to define groups of networks and hosts. You can
234 refer to them with `+name` in the firewall rules' `source` and `dest`
237 The following example allows HTTP traffic from the `management` IP
240 IN HTTP(ACCEPT) -source +management
242 Standard IP set `management`
243 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
245 This IP set applies only to host firewalls (not VM firewalls). Those
246 ips are allowed to do normal management tasks (PVE GUI, VNC, SPICE,
249 The local cluster network is automatically added to this IP set (alias
250 `cluster_network`), to enable inter-host cluster
251 communication. (multicast,ssh,...)
254 # /etc/pve/firewall/cluster.fw
261 Standard IP set 'blacklist'
262 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
264 Traffic from these ips is dropped by every host's and VM's firewall.
267 # /etc/pve/firewall/cluster.fw
275 Standard IP set 'ipfilter-net*'
276 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
278 These filters belong to a VM's network interface and are mainly used to prevent
279 IP spoofing. If such a set exists for an interface then any outgoing traffic
280 with a source IP not matching its interface's corresponding ipfilter set will
283 For containers with configured IP addresses these sets, if they exist (or are
284 activated via the general `IP Filter` option in the VM's firewall's 'options'
285 tab), implicitly contain the associated IP addresses.
287 For both virtual machines and containers they also implicitly contain the
288 standard MAC-derived IPv6 link-local address in order to allow the neighbor
289 discovery protocol to work.
292 /etc/pve/firewall/<VMID>.fw
294 [IPSET ipfilter-net0] # only allow specified IPs on net0
298 Services and Commands
299 ~~~~~~~~~~~~~~~~~~~~~
301 The firewall runs two service daemons on each node:
303 * pvefw-logger: NFLOG daemon (ulogd replacement).
304 * pve-firewall: updates iptables rules
306 There is also a CLI command named 'pve-firewall', which can be used to
307 start and stop the firewall service:
312 To get the status use:
314 # pve-firewall status
316 The above command reads and compiles all firewall rules, so you will
317 see warnings if your firewall configuration contains any errors.
319 If you want to see the generated iptables rules you can use:
329 FTP is an old style protocol which uses port 21 and several other dynamic ports. So you
330 need a rule to accept port 21. In addition, you need to load the 'ip_conntrack_ftp' module.
333 modprobe ip_conntrack_ftp
335 and add `ip_conntrack_ftp` to '/etc/modules' (so that it works after a reboot) .
337 Suricata IPS integration
338 ^^^^^^^^^^^^^^^^^^^^^^^^
340 If you want to use the http://suricata-ids.org/[Suricata IPS]
341 (Intrusion Prevention System), it's possible.
343 Packets will be forwarded to the IPS only after the firewall ACCEPTed
346 Rejected/Dropped firewall packets don't go to the IPS.
348 Install suricata on proxmox host:
351 # apt-get install suricata
352 # modprobe nfnetlink_queue
355 Don't forget to add `nfnetlink_queue` to '/etc/modules' for next reboot.
357 Then, enable IPS for a specific VM with:
360 # /etc/pve/firewall/<VMID>.fw
367 `ips_queues` will bind a specific cpu queue for this VM.
369 Available queues are defined in
372 # /etc/default/suricata
378 include::copyright.adoc[]
384 The firewall contains a few IPv6 specific options. One thing to note is that
385 IPv6 does not use the ARP protocol anymore, and instead uses NDP (Neighbor
386 Discovery Protocol) which works on IP level and thus needs IP addresses to
387 succeed. For this purpose link-local addresses derived from the interface's MAC
388 address are used. By default the 'NDP' option is enabled on both host and VM
389 level to allow neighbor discovery (NDP) packets to be sent and received.
391 Beside neighbor discovery NDP is also used for a couple of other things, like
392 autoconfiguration and advertising routers.
394 By default VMs are allowed to send out router solicitation messages (to query
395 for a router), and to receive router advetisement packets. This allows them to
396 use stateless auto configuration. On the other hand VMs cannot advertise
397 themselves as routers unless the 'Allow Router Advertisement' (`radv: 1`) option
400 As for the link local addresses required for NDP, there's also an 'IP Filter'
401 (`ipfilter: 1`) option which can be enabled which has the same effect as adding
402 an `ipfilter-net*` ipset for each of the VM's network interfaces containing the
403 corresponding link local addresses. (See the
404 <<ipfilter-section,Standard IP set 'ipfilter-net*'>> section for details.)
406 Avoiding link-local addresses on tap and veth devices
407 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
409 With IPv6 enabled by default every interface gets a MAC-derived link local
410 address. However, most devices on a typical {pve} setup are connected to a
411 bridge and so the bridge is the only interface which really needs one.
413 To disable a link local address on an interface you can set the interface's
414 `disable_ipv6` sysconf variable. Despite the name, this does not prevent IPv6
415 traffic from passing through the interface when routing or bridging, so the
416 only noticeable effect will be the removal of the link local address.
418 The easiest method of achieving this setting for all newly started VMs is to
419 set it for the `default` interface configuration and enabling it explicitly on
420 the interfaces which need it. This is also the case for other settings such as
421 `forwarding`, `accept_ra` or `autoconf`.
423 Here's a possible setup:
425 # /etc/sysconf.d/90-ipv6.conf
427 net.ipv6.conf.default.forwarding = 0
428 net.ipv6.conf.default.proxy_ndp = 0
429 net.ipv6.conf.default.autoconf = 0
430 net.ipv6.conf.default.disable_ipv6 = 1
431 net.ipv6.conf.default.accept_ra = 0
433 net.ipv6.conf.lo.disable_ipv6 = 0
437 # /etc/network/interfaces
439 iface vmbr0 inet6 static
444 pre-up echo 0 > /proc/sys/net/ipv6/conf/$IFACE/disable_ipv6