4 include::attributes.txt[]
9 pve-firewall - PVE Firewall Daemon
15 include::pve-firewall.8-synopsis.adoc[]
25 include::attributes.txt[]
28 Proxmox VE Firewall provides an easy way to protect your IT
29 infrastructure. You can setup firewall rules for all hosts
30 inside a cluster, or define rules for virtual machines and
31 containers. Features like firewall macros, security groups, IP sets
32 and aliases helps to make that task easier.
34 While all configuration is stored on the cluster file system, the
35 iptables based firewall runs on each cluster node, and thus provides
36 full isolation between virtual machines. The distributed nature of
37 this system also provides much higher bandwidth than a central
40 The firewall has full support for IPv4 and IPv6. IPv6 support is fully
41 transparent, and we filter traffic for both protocols by default. So
42 there is no need to maintain a different set of rules for IPv6.
48 The Proxmox VE firewall groups the network into the following logical zones:
52 Traffic from/to a cluster node
56 Traffic from/to a specific VM
58 For each zone, you can define firewall rules for incoming and/or
65 All firewall related configuration is stored on the proxmox cluster
66 file system. So those files are automatically distributed to all
67 cluster nodes, and the 'pve-firewall' service updates the underlying
68 iptables rules automatically on changes.
70 You can configure anything using the GUI (i.e. Datacenter -> Firewall,
71 or on a Node -> Firewall), or you can edit the configuration files
72 directly using your preferred editor.
74 Firewall configuration files contains sections of key-value
75 pairs. Lines beginning with a '#' and blank lines are considered
76 comments. Sections starts with a header line containing the section
77 name enclosed in '[' and ']'.
83 The cluster wide firewall configuration is stored at:
85 /etc/pve/firewall/cluster.fw
87 The configuration can contain the following sections:
91 This is used to set cluster wide firewall options.
93 include::pve-firewall-cluster-opts.adoc[]
97 This sections contains cluster wide firewall rules for all nodes.
101 Cluster wide IP set definitions.
105 Cluster wide security group definitions.
109 Cluster wide Alias definitions.
112 Enabling the Firewall
113 ^^^^^^^^^^^^^^^^^^^^^
115 The firewall is completely disabled by default, so you need to
116 set the enable option here:
120 # enable firewall (cluster wide setting, default is disabled)
124 IMPORTANT: If you enable the firewall, traffic to all hosts is blocked by
125 default. Only exceptions is WebGUI(8006) and ssh(22) from your local
128 If you want to administrate your {pve} hosts from remote, you
129 need to create rules to allow traffic from those remote IPs to the web
130 GUI (port 8006). You may also want to allow ssh (port 22), and maybe
133 TIP: Please open a SSH connection to one of your {PVE} hosts before
134 enabling the firewall. That way you still have access to the host if
135 something goes wrong .
137 To simplify that task, you can instead create an IPSet called
138 'management', and add all remote IPs there. This creates all required
139 firewall rules to access the GUI from remote.
142 Host specific Configuration
143 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
145 Host related configuration is read from:
147 /etc/pve/nodes/<nodename>/host.fw
149 This is useful if you want to overwrite rules from 'cluster.fw'
150 config. You can also increase log verbosity, and set netfilter related
151 options. The configuration can contain the following sections:
155 This is used to set host related firewall options.
157 include::pve-firewall-host-opts.adoc[]
161 This sections contains host specific firewall rules.
164 VM/Container configuration
165 ~~~~~~~~~~~~~~~~~~~~~~~~~~
167 VM firewall configuration is read from:
169 /etc/pve/firewall/<VMID>.fw
171 and contains the following data:
175 This is used to set VM/Container related firewall options.
177 include::pve-firewall-vm-opts.adoc[]
181 This sections contains VM/Container firewall rules.
189 IP Alias definitions.
192 Enabling the Firewall for VMs and Containers
193 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
195 Each virtual network device has its own firewall enable flag. So you
196 can selectively enable the firewall for each interface. This is
197 required in addition to the general firewall 'enable' option.
199 The firewall requires a special network device setup, so you need to
200 restart the VM/container after enabling the firewall on a network
207 Firewall rules consists of a direction (`IN` or `OUT`) and an
208 action (`ACCEPT`, `DENY`, `REJECT`). You can also specify a macro
209 name. Macros contain predifined sets of rules and options. Rules can be disabled by prefixing them with '|'.
211 .Firewall rules syntax
215 DIRECTION ACTION [OPTIONS]
216 |DIRECTION ACTION [OPTIONS] # disabled rule
218 DIRECTION MACRO(ACTION) [OPTIONS] # use predefined macro
221 The following options can be used to refine rule matches.
223 include::pve-firewall-rules-opts.adoc[]
225 Here are some examples:
229 IN SSH(ACCEPT) -i net0
230 IN SSH(ACCEPT) -i net0 # a comment
231 IN SSH(ACCEPT) -i net0 -source 192.168.2.192 # only allow SSH from 192.168.2.192
232 IN SSH(ACCEPT) -i net0 -source 10.0.0.1-10.0.0.10 # accept SSH for ip range
233 IN SSH(ACCEPT) -i net0 -source 10.0.0.1,10.0.0.2,10.0.0.3 #accept ssh for ip list
234 IN SSH(ACCEPT) -i net0 -source +mynetgroup # accept ssh for ipset mynetgroup
235 IN SSH(ACCEPT) -i net0 -source myserveralias #accept ssh for alias myserveralias
237 |IN SSH(ACCEPT) -i net0 # disabled rule
239 IN DROP # drop all incoming packages
240 OUT ACCEPT # accept all outgoing packages
246 A security group is a collection of rules, defined at cluster level, which
247 can be used in all VMs' rules. For example you can define a group named
248 `webserver` with rules to open the http and https ports.
251 # /etc/pve/firewall/cluster.fw
254 IN ACCEPT -p tcp -dport 80
255 IN ACCEPT -p tcp -dport 443
258 Then, you can add this group to a VM's firewall
261 # /etc/pve/firewall/<VMID>.fw
271 IP Aliases allow you to associate IP addresses of networks with a
272 name. You can then refer to those names:
274 * inside IP set definitions
275 * in `source` and `dest` properties of firewall rules
277 Standard IP alias `local_network`
278 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
280 This alias is automatically defined. Please use the following command
281 to see assigned values:
284 # pve-firewall localnet
285 local hostname: example
286 local IP address: 192.168.2.100
287 network auto detect: 192.168.0.0/20
288 using detected local_network: 192.168.0.0/20
291 The firewall automatically sets up rules to allow everything needed
292 for cluster communication (corosync, API, SSH) using this alias.
294 The user can overwrite these values in the cluster.fw alias
295 section. If you use a single host on a public network, it is better to
296 explicitly assign the local IP address
299 # /etc/pve/firewall/cluster.fw
301 local_network 1.2.3.4 # use the single ip address
307 IP sets can be used to define groups of networks and hosts. You can
308 refer to them with `+name` in the firewall rules' `source` and `dest`
311 The following example allows HTTP traffic from the `management` IP
314 IN HTTP(ACCEPT) -source +management
316 Standard IP set `management`
317 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
319 This IP set applies only to host firewalls (not VM firewalls). Those
320 ips are allowed to do normal management tasks (PVE GUI, VNC, SPICE,
323 The local cluster network is automatically added to this IP set (alias
324 `cluster_network`), to enable inter-host cluster
325 communication. (multicast,ssh,...)
328 # /etc/pve/firewall/cluster.fw
335 Standard IP set 'blacklist'
336 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
338 Traffic from these ips is dropped by every host's and VM's firewall.
341 # /etc/pve/firewall/cluster.fw
349 Standard IP set 'ipfilter-net*'
350 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
352 These filters belong to a VM's network interface and are mainly used to prevent
353 IP spoofing. If such a set exists for an interface then any outgoing traffic
354 with a source IP not matching its interface's corresponding ipfilter set will
357 For containers with configured IP addresses these sets, if they exist (or are
358 activated via the general `IP Filter` option in the VM's firewall's 'options'
359 tab), implicitly contain the associated IP addresses.
361 For both virtual machines and containers they also implicitly contain the
362 standard MAC-derived IPv6 link-local address in order to allow the neighbor
363 discovery protocol to work.
366 /etc/pve/firewall/<VMID>.fw
368 [IPSET ipfilter-net0] # only allow specified IPs on net0
373 Services and Commands
374 ---------------------
376 The firewall runs two service daemons on each node:
378 * pvefw-logger: NFLOG daemon (ulogd replacement).
379 * pve-firewall: updates iptables rules
381 There is also a CLI command named 'pve-firewall', which can be used to
382 start and stop the firewall service:
387 To get the status use:
389 # pve-firewall status
391 The above command reads and compiles all firewall rules, so you will
392 see warnings if your firewall configuration contains any errors.
394 If you want to see the generated iptables rules you can use:
405 FTP is an old style protocol which uses port 21 and several other dynamic ports. So you
406 need a rule to accept port 21. In addition, you need to load the 'ip_conntrack_ftp' module.
409 modprobe ip_conntrack_ftp
411 and add `ip_conntrack_ftp` to '/etc/modules' (so that it works after a reboot) .
414 Suricata IPS integration
415 ~~~~~~~~~~~~~~~~~~~~~~~~
417 If you want to use the http://suricata-ids.org/[Suricata IPS]
418 (Intrusion Prevention System), it's possible.
420 Packets will be forwarded to the IPS only after the firewall ACCEPTed
423 Rejected/Dropped firewall packets don't go to the IPS.
425 Install suricata on proxmox host:
428 # apt-get install suricata
429 # modprobe nfnetlink_queue
432 Don't forget to add `nfnetlink_queue` to '/etc/modules' for next reboot.
434 Then, enable IPS for a specific VM with:
437 # /etc/pve/firewall/<VMID>.fw
444 `ips_queues` will bind a specific cpu queue for this VM.
446 Available queues are defined in
449 # /etc/default/suricata
453 Avoiding link-local addresses on tap and veth devices
454 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
456 With IPv6 enabled by default every interface gets a MAC-derived link local
457 address. However, most devices on a typical {pve} setup are connected to a
458 bridge and so the bridge is the only interface which really needs one.
460 To disable a link local address on an interface you can set the interface's
461 `disable_ipv6` sysconf variable. Despite the name, this does not prevent IPv6
462 traffic from passing through the interface when routing or bridging, so the
463 only noticeable effect will be the removal of the link local address.
465 The easiest method of achieving this setting for all newly started VMs is to
466 set it for the `default` interface configuration and enabling it explicitly on
467 the interfaces which need it. This is also the case for other settings such as
468 `forwarding`, `accept_ra` or `autoconf`.
470 Here's a possible setup:
472 # /etc/sysconf.d/90-ipv6.conf
474 net.ipv6.conf.default.forwarding = 0
475 net.ipv6.conf.default.proxy_ndp = 0
476 net.ipv6.conf.default.autoconf = 0
477 net.ipv6.conf.default.disable_ipv6 = 1
478 net.ipv6.conf.default.accept_ra = 0
480 net.ipv6.conf.lo.disable_ipv6 = 0
484 # /etc/network/interfaces
487 iface vmbr0 inet static
489 netmask 255.255.255.128
491 iface vmbr0 inet6 static
496 pre-up echo 0 > /proc/sys/net/ipv6/conf/$IFACE/disable_ipv6
498 # With IPv6-only 'pre-up' is too early and 'up' is too late.
499 # Work around this by creating the bridge manually
500 iface vmbr1 inet manual
501 pre-up ip link add $IFACE type bridge
502 up echo 0 > /proc/sys/net/ipv6/conf/$IFACE/disable_ipv6
503 iface vmbr1 inet6 static
509 bridge_vlan_aware yes
518 The firewall contains a few IPv6 specific options. One thing to note is that
519 IPv6 does not use the ARP protocol anymore, and instead uses NDP (Neighbor
520 Discovery Protocol) which works on IP level and thus needs IP addresses to
521 succeed. For this purpose link-local addresses derived from the interface's MAC
522 address are used. By default the 'NDP' option is enabled on both host and VM
523 level to allow neighbor discovery (NDP) packets to be sent and received.
525 Beside neighbor discovery NDP is also used for a couple of other things, like
526 autoconfiguration and advertising routers.
528 By default VMs are allowed to send out router solicitation messages (to query
529 for a router), and to receive router advetisement packets. This allows them to
530 use stateless auto configuration. On the other hand VMs cannot advertise
531 themselves as routers unless the 'Allow Router Advertisement' (`radv: 1`) option
534 As for the link local addresses required for NDP, there's also an 'IP Filter'
535 (`ipfilter: 1`) option which can be enabled which has the same effect as adding
536 an `ipfilter-net*` ipset for each of the VM's network interfaces containing the
537 corresponding link local addresses. (See the
538 <<ipfilter-section,Standard IP set 'ipfilter-net*'>> section for details.)
541 Ports used by Proxmox VE
542 ------------------------
544 * Web interface: 8006
545 * VNC Web console: 5900-5999
547 * sshd (used for cluster actions): 22
549 * corosync multicast (if you run a cluster): 5404, 5405 UDP
557 include::pve-firewall-macros.adoc[]
560 include::pve-copyright.adoc[]