PfBlocker

LinuxTracker

Came across a new list of malware IPs

http://labs.alienvault.com/labs/index.php/projects/open-source-ip-reputation-portal/download-ip-reputation-database/

Looks like either the gzip ver or the plain ver - of the Generic Format - would load fine in pfBlocker.

Haven't tried it yet - will give it a go tonight and post back how it loads.

Here's my report back on the AlienVault IP Reputation list.
I had it block matching http/https outbound - and block matching * inbound.

The list is big, over 300k IPs.

The text version is 17MB and it added nearly a minute to my filter reload time (3Ghz PentiumD, 4GB).

I tried the gzip version and found it much smaller and faster. I had no noticeable difference in filter reload times.

After running the list for a most of a week, I had 2 false positives. That's not bad.

One was a big web host that AlienVault listed over a year ago (for serious malware distribution).
The IP was cleaned immediately so AlienVault must not delist IPs automatically.
The other false positive was a Tor directory server.

Conclusion:
Given the big database (and no auto-delisting) I'd say AlienVault has a really low false positive rate.
My opinion is that it's worth trying - providing you log any hits on it.

LinuxTracker

@AudiAddict:

Some feature requests, which I hope you can add to make it even better:

1) Traffic being blocked by pfblocker logged in a different tab (instead of the firewall log?) Easy to debug if something is being blocked by my own rules or the pfblocker app.
I do know that you can see that if something is blocked by pfblocker you can see that by looking at the details, but when you have allot of blocked records it's hard to filter

2) Users are complaining specific sites are being blocked. Is there a whitelisting feature available to ignore certain websites? Or do I have to enter them in a alias in my firewall manually?

5) Ignore tab –> where I can enter my own ip for outbound and my servers for inbound? Just to make 100 % sure these do not get blocked?

#1 - I'd sort of like that. Prob is I have 10+ block lists in addition to pfBlocker's country lists. I'd want them all logged individually and that doesn't seem practical.
Maybe a filtered syslog server could be a solution.

For now, looking at ports/IPs in the firewall log can clue me in on what list is blocking.

If I suspect a site is blocked and I need to figure out which list is responsible; I'll run something like```
grep -r -o -Dskip 111.222.3.4 /


#2 I have a Whitelist alias that I keep at the top of the rules list.  I figure adding an IP to the alias is as simple as any option pfBlocker could provide.
**Edit: You could also create a new Custom List in pfBlocker and set it to the top of your rules list.**

#5 - See #2 (providing I understand correctly).

LinuxTracker

@caustic386:

I have some rules that are being autocreated twice. Any timeline available?

I have that happen occasionally.
I'll delete the dupes and finish my usual setup and it doesn't bother me again after that.

Just for the record, my usual setup:

Under Rules: Edit - I make sure the "Auto Rule" text is chopped off the end of the Description - of the rules I want to keep.
I delete the duplicate rules (and reorder the remaining ones if necessary). Apply rule changes.
Under pfBlocker -> Lists - I make sure Action is set to Alias_only for every list and repeat for Continent tabs.

Unless Action is set to Alias_only, pfBlocker will:
- Create a new rule if the old one is modified
- Restore the original order of it's rules

LinuxTracker

For anyone who's read the I-Blocklist doesn't protect Bittorrent users from being monitored article;
http://torrentfreak.com/anti-pirates-caught-spying-on-thousands-of-torrents-120829/
I put together a block list that comprises all 6 monitoring subnets, mentioned in the original study.

http://dl.dropbox.com/u/71477228/P2PCIDRList.txt

Original P2P monitoring Study:
http://www.cs.bham.ac.uk/~tpc/Papers/P2PSecComm2012.pdf
If you P2P, I strongly recommend you read it.

A more accurate summary of the study is that I-Blocklist missed about a third of the monitoring IPs and also had a lot of false positives.
They study authors mentioned 6 subnets that harbored P2P monitors and that's what's in this list.

Disclaimer: My list is serious overkill. 99.9% of the IP addresses in it don't monitor anything.
The list is what it is because the study authors didn't mention specific IPs, just AS#'s (page #14).
So the list is all the IPs that fall under those AS#'s.

LinuxTracker

@LinuxTracker:

@caustic386:

I have some rules that are being autocreated twice. Any timeline available?

I have that happen occasionally.
I'll delete the dupes and finish my usual setup and it doesn't bother me again after that.

On 2nd thought, maybe I shouldn't dismiss caustic386's point so quickly.

I added a list yesterday and had 5 duplicate rules (6 total) show up in WAN.

My theory is that it's tied to the length of time it takes my firewall rules to reload.
Could there be a loop that repeats until the firewall rules finish loading?

(I have a lot of lists in my pfSense rig. I suspect if I add more lists, I'll have a greater # of dupes created.)

All my pfBlocker lists are configured as Alias - so not a big deal for me.
But this could be a real issue for someone who set their pfBlocker list config as Deny or Allow. Dupe rules could be created at every list reload.

LinuxTracker

Here are Spamhaus criminal network block lists.
Drop: http://www.spamhaus.org/drop/drop.txt
EDrop: http://www.spamhaus.org/drop/edrop.txt

DROP (Don't Route Or Peer) and EDROP are advisory "drop all traffic" lists, consisting of stolen 'hijacked' netblocks and netblocks controlled entirely by criminals and professional spammers.

The DROP list will not include any IP address space under the control of any legitimate network - even if being used by "the spammers from hell".
DROP will only include netblocks allocated directly by an established Regional Internet Registry (RIR) or National Internet Registry (NIR) such as ARIN, RIPE, AFRINIC, APNIC, LACNIC or KRNIC
or direct RIR allocations illicitly taken from the original allocatee, that is, the troubling run of "hijacked" IP address blocks that have been snatched away from their original owners (which in most cases are long dead corporations) and are now controlled by netblock thieves or sold to criminal spammers.

EDROP is an extension of the DROP list that includes suballocated netblocks controlled by cyber criminals.
EDROP is meant to be used in addition to the direct allocations on the DROP list.

From http://www.spamhaus.org/drop/

marcelloc

@LinuxTracker:

Here are Spamhaus criminal network block lists.

Thank's LinuxTracker. These lists are really usefull. :)

Rezin

@LinuxTracker:

I have that happen occasionally.

Legend. Thank you. Fixed the dupe issue for me, and allowed a rule (that I'd setup to bypass the pfBlocker rule) to stay before the pfBlocker rule.

LinuxTracker

I made a list so I can block spam from Tor Exit nodes.
http://dl.dropbox.com/u/71477228/torexitnodes.txt

It pulls the IPs from here -> http://exitlist.torproject.org/ and updates every hour.

There are a lot of Tor Node lists, even a number of Tor Exit Node lists.
EmergingThreats, TorProject and dan.me.uk all maintain them

For whatever reason, I couldn't find a list of just Exit Nodes in the right format for pfBlocker.
I found about everything else though.

As example, torexit.dan.me.uk is a DNSBL w/ a return value of 127.0.0.100
re: https://www.dan.me.uk/dnsbl

And this guy offers a Bash Script that will pull from dan.me.uk and upload right to IPTables.
https://github.com/jseidl/torblock/blob/master/torblock.sh

But I couldn't find exactly what I wanted so I made it and there you are.

phil.davis

@caustic386:

I have some rules that are being autocreated twice. Any timeline available?

I found out how to make this happen - On the GUI take a rule in the middle of your WAN rules, move it to the very end of the list. When you apply, the rule ends up by itself at the very end of the list of rules in the filter section of config.xml - there are WAN rules, then LAN rules then this 1 WAN rule. If you have lots of interfaces and move your rules order around from time to time, the rules for each interface end up in interspersed chunks in config.xml. Everything is fine for correct rule processing for pf, the rules for each interface are in the correct order in the filter section, it is just that rules for WAN, LAN, OPT1 OpenVPN etc get interspersed.
The pfBlocker code would apply pfBlocker rules every time it came to a "chunk" of rules for another interface. If there were 2 "chunks" of WAN rules in the filter section, then pfBlocker rules would be added before both "chunks".
I have fixed the logic so pfBlocker only adds the rules the very first time it comes across an interface.
@marcelloc - have a look at pull request https://github.com/bsdperimeter/pfsense-packages/pull/323 - it should fix this issue for good.

marcelloc

@phil.davis:

@marcelloc - have a look at pull request https://github.com/bsdperimeter/pfsense-packages/pull/323 - it should fix this issue for good.

Thank's phil.davis, pull request merged. :)

Cino

@marcelloc:

@phil.davis:

@marcelloc - have a look at pull request https://github.com/bsdperimeter/pfsense-packages/pull/323 - it should fix this issue for good.

Thank's phil.davis, pull request merged. :)

I just finish an update to 2.1-BETA0 (i386) built on Mon Oct 1 14:51:39 EDT 2012. After all my packages were re-installed; I noticed there were no rules created by pfBlocker.. Uninstalled, rebooted then installed the package again.. It still wouldn't create the rules..

I reverted the change to https://github.com/bsdperimeter/pfsense-packages/blob/77d63137cfaafe5786ca958833558dcf0837a145/config/pf-blocker/pfblocker.inc and the rules were auto created in the interfaces.. Could this be an issue just with 2.1 and pfBlocker?

phil.davis

@Cino - you quoted some text from marcelloc and phil.davis (me), but your link to a commit of pfblocker.inc is a commit from a few months ago. I stared at the latest commit again at https://github.com/bsdperimeter/pfsense-packages/commit/90acaa6951653a8c40a4a14f9f0629e74ba47b81 but can't see what would go wrong.
Exactly which change did you revert to make it go?

phil.davis

Sorry, I looked harder, there was an "=" where there should have been "==" in a test in /usr/local/www/pfblocker.inc:

-          if ($processed_interface = $rule['interface']){
+          if ($processed_interface == $rule['interface']){

Add the extra "=" and it works! Pull request https://github.com/bsdperimeter/pfsense-packages/pull/325/files

marcelloc

Fix merged. :)

Cino

that did the trick! Thanks for the quick fix Phil!!

tommyboy180

Big difference between == and =

Nice catch.

fuzzyozzy408

Hi,

My background is cisco networking. i am trying to find a way to script the enabling / disabling of pfBlocker. This is for shutdown / startup purposes. i have a cron job to shut the server down in the morning and start it up again at night (i work during the day and summer days are very hot where i live).

Unfortunately it does not seem obvious to me, how to enable/disable pfBlocker by script. I have got as far as reading the /usr/local/pkg/pfblocker.xml + .inc files - but i am not knowledge with php. Could someone please advise how i can got about scripting the enabling/disabling of pfBlocker?

Thanks in advance
dirk

PS: This is my first post and i hope i am posting in the right area. Please redirect me if i have posted in the wrong place.

marcelloc

follow these steps:
rename current pfblocker applied rules description
change pfblocker action to alias only
create a schedule
apply it to pfblocker rules you renamed descriptions

fuzzyozzy408

Thanks Marcelloc,

I should have given more detail. For whatever reasons, snort does not start up, if i reboot the server with pfblocker enabled. Its an error i have tried to resolve and failed. (i am guessing i have been overenthusiastic subscribing to blacklists and blocking many countries as well.).

In the cron'ed script i would like to disable pfblocker and then halt the server. If the server is booted, i would like to enable pfblocker via one of the RC scripts. Obviously i need a largish sleep count, to make sure snort has started up first.