Had my pfSense been compromised?

ASIC

When I place the mouse over the green check on left on the rule, it shows "pass/0". When I click it opens a popup: "The rule that triggered this action is:" Nothing else.

ASIC

Rule description is @4294967295.

hulleyrob

Does anyone have an idea where the rule id comes from as it looks like we all have the same rule ID number creating these rules maybe this will give us an idea of where they are coming from?

Ideally there would be some logging or history but as everyone has said upnp is a bad idea for most people so i guess thats why it doesnt exist.

Also they are also 8 seconds apart in batches but this could be the time that the service runs each time?

johnpoz

So I just turned UPnP on - to test.. And then set my plex to use UPnP vs the static forward I have setup.

So I disabled my normal port forward.. Then enable UPnP

And I get this.

After doing a can you see me for the port listed in the UPnP status 21735

Is not coming up with that ID number.. A google for that ID and pfsense came up with this thread
https://forum.netgate.com/topic/128685/firewall-log-showing-strange-pass-entry

Can we see the raw firewall logs for these entries. Like in that thread.

example - here is the raw entry for one of those passed via upnp

Oct 15 06:10:27 	filterlog: 224,0,miniupnpd,0,igb1,match,pass,in,4,0x0,,42,13284,0,DF,6,tcp,60,52.202.215.126,192.168.9.10,58368,32400,0,S,2623792563,,26883,,mss;sackOK;TS;nop;wscale

You can see from the log that miniupnpd there.. So lets see the raw for these entries.

From that other thread
Mar 25 14:42:05 pfSense filterlog: 4294967295,,,0,re0,short,pass,in,4,0x0,,239,10359,0,DF,6,tcp,36,105.212.87.78,[MYIPADDRESS],2123,23,-4,S,errormsg='[bad hdr length 20 - too long, > 16]',

Not exactly sure what "short" means as the reason...

hulleyrob

@johnpoz yeah sure i did look at those last night but couldnt see anything that looked much differant from in the screenshot but maybe you will see something i cannot. Will post as soon as i get home from work. Thanks to everyone for input and testing this.

ASIC

Here is my raw filter log for this:
Oct 14 20:42:00 filterlog: 4294967295,,,0,pppoe0,short,pass,in,4,0x0,,241,53405,0,none,17,udp,27,103.240.140.10,[MYIPADDRESS],1810,1985,7

johnpoz

@ASIC said in Had my pfSense been compromised?:

short,

I have no idea what that means... Have never seen that in the logs before.

Found this;
reason res - the reason that action was taken. Possible reasons are match, bad-offset, fragment, short, normalize, memory, bad-timestamp, congestion, ip-option, proto-cksum, state-mismatch, state-insert, state-limit, src-limit and synproxy.

But not sure what "short" actually means?

edit: Ok looks there have been some that match on short (2) - but still not actually finding anything that describes what it actually means

[2.4.4-RELEASE][admin@sg4860.local.lan]/root: pfctl -s info
Status: Enabled for 30 days 22:17:00          Debug: Urgent

Interface Stats for igb0              IPv4             IPv6
  Bytes In                    350478233746         69232705
  Bytes Out                  1405455628533       1128183040
  Packets In
    Passed                       466867557           311496
    Blocked                           8876                1
  Packets Out
    Passed                       999590646          1171599
    Blocked                              0                0

State Table                          Total             Rate
  current entries                      404               
  searches                      3535810440         1323.2/s
  inserts                         15778842            5.9/s
  removals                        15778438            5.9/s
Counters
  match                           16542390            6.2/s
  bad-offset                             0            0.0/s
  fragment                              64            0.0/s
  short                                  2            0.0/s
  normalize                             21            0.0/s
  memory                                 0            0.0/s
  bad-timestamp                          0            0.0/s
  congestion                             0            0.0/s
  ip-option                            125            0.0/s
  proto-cksum                            0            0.0/s
  state-mismatch                      2834            0.0/s
  state-insert                           0            0.0/s
  state-limit                            0            0.0/s
  src-limit                              0            0.0/s
  synproxy                               0            0.0/s
  map-failed                             0            0.0/s
[2.4.4-RELEASE][admin@sg4860.local.lan]/root:

bchan

@ASIC said in Had my pfSense been compromised?:

Rule description is @4294967295.

Same id as mine.

ASIC

@bchan said in Had my pfSense been compromised?:

4294967295

Maybe because 4294967295(Dec) is equal to FFFFFFFF(Hex).

johnpoz

Ok so from research so fare

short packets dropped

Is that that is suppose to mean.. but why does it say passed then?

provels

So there is no outbound traffic to 103.240.140.10 from some device that creates this return Allow temp rule? Since all return traffic is allowed? Create a rule that allows (but logs) all traffic to 103.240.140.10? (LOL, ignore as needed! :)

johnpoz

To be honest I think this is log of traffic that IP or transport header is too short.. But not sure why its saying pass... But I don't think that traffic is actually being passed to anything, not even pfsense.

hulleyrob

Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35606,0,none,17,udp,27,103.240.140.10,IP_REMOVED,3266,990,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35608,0,none,17,udp,27,103.240.140.10,IP_REMOVED,2427,389,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35605,0,none,17,udp,27,103.240.140.10,IP_REMOVED,1862,1025,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35607,0,none,17,udp,27,103.240.140.10,IP_REMOVED,570,995,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35604,0,none,17,udp,27,103.240.140.10,IP_REMOVED,493,514,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35609,0,none,17,udp,27,103.240.140.10,IP_REMOVED,2375,445,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35602,0,none,17,udp,27,103.240.140.10,IP_REMOVED,162,546,7
Oct 13 09:14:13 pfSense filterlog: 4294967295,16777216,,0,pppoe0,short,pass,in,4,0x0,,245,35603,0,none,17,udp,27,103.240.140.10,IP_REMOVED,3871,547,7

Here are the entries from my log file.

johnpoz

So it seems these are short, ie something wrong with them... But the question is why are they getting logged as passed.. If they are short - they should just be dropped.

hulleyrob

@johnpoz if thats all it is it will be a relief as you saw on the screenshot above it looked like something had opened up all those ports.

What do I do now, is it worth reporting this somewhere for investigation?

johnpoz

I would hope @jimp or @Derelict might be able to give more details of these "short" log entries..

jimp

I can't remember the last time I saw one marked short. It might be a fragment.

johnpoz

But the question is why is marked pass?

jimp

If it's a fragment that is part of an existing connection, it would be passed. It might be that the state recently expired or was purged for some other reason so it was passed in but NAT didn't get applied, for example.

johnpoz

I doubt its frag, since that is a different counter - see my output from
pfctl -s info

above.

From the logs he lists I doubt its part of an existing connection.. From everything I find short should be dropped, so why is it logged as if it passed?