How to read status output

From Libreswan
Jump to navigation Jump to search

The libreswan status output is very verbose and confusing. Work is being done to make this a lot more userfriendly.

There are currently two status commands that can be used. Note all status commands prefix their output using "FTP status codes" in the form of three digits (eg 000 or 2xx or 5xx)


ipsec trafficstatus

The "ipsec whack --trafficstatus" command shows the tunnels that are currently established. An example output: 

006 #48971: "redhat", username=pwouters, type=ESP, add_time=1480687195, inBytes=12295525, outBytes=673668
006 #50578: "supo.libreswan.org", type=ESP, add_time=1480701786, inBytes=0, outBytes=0, id='@melto.foobar.fi'
006 #50206: "vault.libreswan.org", type=ESP, add_time=1480698400, inBytes=1008, outBytes=1008, id='@foo-gw.foobar.fi'

This output shows that there are currently three connections established. The first connection is named "redhat". It has a username (XAUTH or IKEv2 CP) associated with it of "pwouters". It shows when the connection became active (add_time is the number of seconds since Unix Epoch) and a traffic count for incoming and outgoing traffic. The other two tunnels show simiar information, except that since these connections specified a remote ID to connect to, these IDs are also listed.

This short output does not reveal any details of the connection. It does not show if IKEv1 or IKEv2 was used.

ipsec status

The "ipsec status" command shows a more verbose but not very userfriendly output. This command is extremely verbose and was originally a developer-only tool for debugging. It is not really designed for administrators. Work is underway to replace this output with something more human readable.

The output consists of a few sections seperated by blanc lines:

  • Network interfaces used
  • Global configuration (from ipsec.conf's "config setup"
  • Supported kernel algorithms for ESP/AH
  • Supported userland algorithms for IKE
  • Connection list (loaded connections - not neccessarilly active)
  • State Information (including count of IKE and IPsec states)
  • A list of all active IKE and IPsec states
  • Bare shunts (packet-triggered policies for which no loaded connection matched)

We are interested in the connection list and the state list. Connection parameters are those parameters loaded from the configuration. State parameters are those parameters that are being or were negotiated for this particular negotiation of that connection.


Connection listing example

The following example output is the "redhat" connection as taken from the connection list section: 

000 "redhat": 10.3.228.119/32===192.168.10.124[@OURREMOTEID,+MC+XC+S=C]---192.168.10.1...209.XXX.XXX.XXX<XXX.XXX.redhat.com>[MS+XS+S=C]===0.0.0.0/0; erouted; eroute owner: #2
000 "redhat":     oriented; my_ip=10.3.228.119; their_ip=unset
000 "redhat":   xauth us:client, xauth them:server,  my_username=pwouters; their_username=[any]
000 "redhat":   our auth:secret, their auth:secret
000 "redhat":   modecfg info: us:client, them:server, modecfg policy:push, dns1:unset, dns2:unset, domain:redhat.com, cat:unset;
000 "redhat": banner:Unauthorized Access to this or any other Red Hat Inc. device is strictly prohibited. Violators will be prosecuted. ;
000 "redhat":   labeled_ipsec:no;
000 "redhat":   policy_label:unset;
000 "redhat": 10.3.228.119/32===192.168.10.124[@,+MC+XC+S=C]---192.168.10.1...209.XXX.XXX.XXX[MS+XS+S=C]===10.0.0.0/8; erouted; eroute owner: #2
000 "redhat":     oriented; my_ip=10.3.228.119; their_ip=unset
000 "redhat":   xauth us:client, xauth them:server,  my_username=pwouters; their_username=[any]
000 "redhat":   our auth:secret, their auth:secret
000 "redhat":   modecfg info: us:client, them:server, modecfg policy:push, dns1:unset, dns2:unset, domain:redhat.com, cat:unset;
000 "redhat": banner:Unauthorized Access to this or any other Red Hat Inc. device is strictly prohibited. Violators will be prosecuted. ;
000 "redhat":   labeled_ipsec:no;
000 "redhat":   policy_label:unset;
000 "redhat":   ike_life: 86400s; ipsec_life: 86400s; replay_window: 32; rekey_margin: 540s; rekey_fuzz: 100%; keyingtries: 0;
000 "redhat":   retransmit-interval: 500ms; retransmit-timeout: 60s;
000 "redhat":   sha2-truncbug:yes; initial-contact:yes; cisco-unity:no; fake-strongswan:no; send-vendorid:yes; send-no-esp-tfc:no;
000 "redhat":   policy: PSK+ENCRYPT+TUNNEL+PFS+DONT_REKEY+UP+XAUTH+AGGRESSIVE+IKEV1_ALLOW+IKEV2_ALLOW+IKE_FRAG_ALLOW+NO_IKEPAD+ESN_NO;
000 "redhat":   conn_prio: 32,32; interface: wlp4s0; metric: 0; mtu: unset; sa_prio:666; sa_tfc:none;
000 "redhat":   nflog-group: unset; mark: 6/0xffffffff, 6/0xffffffff; vti-iface:ipsec0; vti-routing:yes; vti-shared:no;
000 "redhat":   dpd: action:hold; delay:30; timeout:60; nat-t: force_encaps:no; nat_keepalive:yes; ikev1_natt:both
000 "redhat":   newest ISAKMP SA: #3; newest IPsec SA: #2;
000 "redhat":   IKE algorithms wanted: AES_CBC(7)_128-SHA1(2)-MODP1024(2)
000 "redhat":   IKE algorithms found:  AES_CBC(7)_128-SHA1(2)-MODP1024(2)
000 "redhat":   IKE algorithm newest: AES_CBC_128-SHA1-MODP1024
000 "redhat":   ESP algorithms wanted: AES(12)_000-SHA1(2); pfsgroup=MODP1024(2)
000 "redhat":   ESP algorithms loaded: AES(12)_000-SHA1(2)
000 "redhat":   ESP algorithm newest: AES_256-HMAC_SHA1; pfsgroup=MODP1024

The most interesting bit is the first line that describes the IP ranges of both ends of the tunnel. The bulk of the output displays the various connection options that are set or unset. The last bit shows the interpretation of the algorithms specified in the connection's ike= and esp= line.

Note this part of the output does not state anything about whether this tunnel is up or down or being negotiatied. That information is visible in the State List:

State Listing

Fully working VPN connections consist of an IKE SA (phase1) and an IPsec SA (phase2). The IKE SA refers to the userland state and describes the endpoints IKE negotiation state. The IPsec SA refers to the kernel state that is actually responsible for encrypting and decrypting the IP packets.

An IKEv1 connection example looks like: 

000 Total IPsec connections: loaded 6, active 1
000
000 State Information: DDoS cookies not required, Accepting new IKE connections
000 IKE SAs: total(1), half-open(0), open(0), authenticated(1), anonymous(0)
000 IPsec SAs: total(1), authenticated(1), anonymous(0)
000
000 #2: "redhat":4500 STATE_QUICK_I2 (sent QI2, IPsec SA established); EVENT_SA_REPLACE_IF_USED in 28039s; newest IPSEC; eroute owner; isakmp#1; idle; import:admin initiate
000 #2: "redhat" esp.9c01f20f@209.132.183.55 esp.e12dd706@192.168.10.124 tun.0@209.132.183.55 tun.0@192.168.10.124 ref=0 refhim=0 Traffic: ESPin=0B ESPout=0B! ESPmax=4194303B username=pwouters
000 #1: "redhat":4500 STATE_MAIN_I4 (ISAKMP SA established); EVENT_SA_EXPIRE in 86390s; newest ISAKMP; lastdpd=-1s(seq in:0 out:0); idle; import:admin initiate
000

In this case there is one connection named "redhat". It has an established IKE SA (also called phase1 or Parent SA) numbered #1 and it has one IPsec SA (also called phase2 or Child SA) numbered #2. Every IPsec SA will list its IKE SA state number as "isakmp#XXX".

If the state claims "established", it means that it is fully up and running.

In IKEv1, the phase1 states are either MAIN MODE or AGGRESSIVE MODE states and the phase2 states are QUICK_MODE states

In IKEv2, the IKE states are called PARENT SA and the IPsec states are called CHILD SA.

A Parent SA (phase1) state that is still negotiating has no Child SA (phase2) state. For example: 

000 #51231: "private-or-clear#193.110.157.0/24"[1] ...193.110.157.102:500 STATE_PARENT_I1 (sent v2I1, expected v2R1); EVENT_v2_RETRANSMIT in 8s; idle; import:local rekey

In this case, we initiated an IKEv2 connection named "private-or-clear" to the remote IP 193.110.157.102, but the remote endpoint never responded to us. It shows the connection will retry in 8 seconds.

An established IKEv2 connection looks as follows:

000 #50578: "supo.libreswan.org":500 STATE_PARENT_R2 (received v2I2, PARENT SA established); EVENT_SA_REPLACE in 22484s; newest IPSEC; eroute owner; isakmp#50577; idle; import:respond to stranger 000 #50578: "supo.libreswan.org" esp.9a8c024f@188.127.201.217 esp.8758b394@76.10.157.69 tun.0@188.127.201.217 tun.0@76.10.157.69 ref=0 refhim=0 Traffic: ESPin=0B ESPout=0B! ESPmax=0B 000 #50577: "supo.libreswan.org":500 STATE_PARENT_R2 (received v2I2, PARENT SA established); EVENT_SA_REPLACE in 22484s; newest ISAKMP; isakmp#0; idle; import:respond to stranger 000 #50577: "supo.libreswan.org" ref=0 refhim=0 Traffic:

Here, state number 50577 is the IKE SA, and state 50578 is the IPsec SA.

The Child SA (#50578) in IKEv2 mistakenly calls itself a a PARENT SA. This is a known (but harmless) bug

ip xfrm

It is also possible to ask the kernel about the IPsec state inside the kernel. The kernel IPsec state consists of two parts. The Security Policy Database (SPD) and the Security Association Database (SAD). In Linux kernel terms these are called "xfrm policy" and "xfrm state". They are linked together by the reqid.

IPsec SA's always come in a bundle. There is an inbound (in) and outbound (out) IPsec SA. They have their own policy that are usually just a mirror image of each other, although each direction as its own encryption and/or authentication keys. There is also a forward (fwd) policy which is an artifact of the implementation and can mostly be ignored. 

# ip xfrm pol
src 10.3.228.211/32 dst 10.0.0.0/8 
	dir out priority 666 ptype main 
	mark 0x6/0xffffffff
	tmpl src 76.10.157.68 dst 209.132.183.55
		proto esp reqid 16417 mode tunnel
src 0.0.0.0/0 dst 10.3.228.211/32 
	dir fwd priority 666 ptype main 
	mark 0x6/0xffffffff
	tmpl src 209.132.183.55 dst 76.10.157.68
		proto esp reqid 16417 mode tunnel
src 0.0.0.0/0 dst 10.3.228.211/32 
	dir in priority 666 ptype main 
	mark 0x6/0xffffffff
	tmpl src 209.132.183.55 dst 76.10.157.68
		proto esp reqid 16417 mode tunnel
src 0.0.0.0/0 dst 0.0.0.0/0 
	socket out priority 0 ptype main 
src 0.0.0.0/0 dst 0.0.0.0/0 
	socket in priority 0 ptype main 
src 0.0.0.0/0 dst 0.0.0.0/0 
	socket out priority 0 ptype main 
src 0.0.0.0/0 dst 0.0.0.0/0 

[...]

The output can look slightly different depending on the kernel version. As libreswan pokes holes for the IKE port (UDP 500) there will be a number of similar looking states to and from 0.0.0.0/0, even if no IPsec connection has been established. The source/destination addresses of the policy determine which packets are send to a kernel state for encryption/decryption. It will match either the libreswan connection's left= and right= IP addresses (for a host-to-host connection) or it will match the libreswan connection leftsubnet= and rightsubnet= values if set. The reqid's from this policy is linked to the states that have the crypto keys:

A more verbose version including traffic counters can be obtained using ip -s xfrm pol

The corresponding xfrm state can be shown using ip xfrm state :

This command actually shows the encryption/decryption and authentication private keys. Anyone who sees this output can decrypt all the traffic encrypted with these keys! 
# ip xfrm state
src 209.132.183.55 dst 76.10.157.68
	proto esp spi 0x6e45ab4b reqid 16417 mode tunnel
	replay-window 32 flag af-unspec
	auth-trunc hmac(sha1) 0x5292bb9d25099ecec68c54457ccd06099b4930ac 96
	enc cbc(aes) 0xd7a05a7fd43ca2110b90e151ec55883c4623f86905178cb9eed5d42e328d31bf
	anti-replay context: seq 0x0, oseq 0x0, bitmap 0x00000000
src 76.10.157.68 dst 209.132.183.55
	proto esp spi 0x360d5bb6 reqid 16417 mode tunnel
	replay-window 32 flag af-unspec
	auth-trunc hmac(sha1) 0x059d0c7ee4a0a1cbbba3efc33aa779abaf8fa0ed 96
	enc cbc(aes) 0x66085742da59ccb01e6c7702675ca6fc7876ede803883ad519ecc805327d97d0
	anti-replay context: seq 0x0, oseq 0x0, bitmap 0x00000000

Here the IP addresses are the IP's of the gateways themselves and not the leftsubnet= or rightsunet= values.

This command also supports the more verbose version using ip -s xfrm state

Retrieved from "https://libreswan.org/wiki/index.php?title=How_to_read_status_output&oldid=22042"