PfBlockerNG v2.0 w/DNSBL

BBcan177

Again, every thing works as expected other than browsing to 10.10.10.1:80 or 10.10.10.1:443 has to time out during page loads. I have looked around and I think I see the rules in place for this but I still think one is wrong or I am missing something simple. Included below are some snips from the config XML

Since you are using multiple VLANS, you need to ensure that all VLANs can hit the DNSBL VIP… There is an option in the DNSBL Tab, to add a Permit Floating rule... Ensure that you select all of the VLANs in the selection menu... You will have slow browsing issues if you can't get ping the DNSBL VIP and Browse to the DNSBL VIP and get the 1x1 pix....

tjbaudio

@BBcan177:

@tjbaudio:

Again, every thing works as expected other than browsing to 10.10.10.1:80 or 10.10.10.1:443 has to time out during page loads. I have looked around and I think I see the rules in place for this but I still think one is wrong or I am missing something simple. Included below are some snips from the config XML

Since you are using multiple VLANS, you need to ensure that all VLANs can hit the DNSBL VIP… There is an option in the DNSBL Tab, to add a Permit Floating rule... Ensure that you select all of the VLANs in the selection menu... You will have slow browsing issues if you can't get ping the DNSBL VIP and Browse to the DNSBL VIP and get the 1x1 pix....

My apologies. That is my ultimate goal. At this point I just have a stock install with 2 nics, WAN and LAN nics. No Vlan's. Just default settings+ pfBNG installed.
Edit to add:
I can ping 10.10.10.1 from LAN.
In states I get:
LAN tcp 10.10.10.1:8081 <- 192.168.1.101:50268 ESTABLISHED:ESTABLISHED 5 / 4 587 B / 576 B
When loading 10.10.10.1:8081 in a browser.

LAN tcp 127.0.0.1:8443 (10.10.10.1:443) <- 192.168.1.135:50120 CLOSED:SYN_SENT 3 / 0 152 B / 0 B
When loading 10.10.10.1:80 in a browser.

LAN tcp 127.0.0.1:8443 (10.10.10.1:443) <- 192.168.1.101:35768 CLOSED:SYN_SENT 2 / 0 120 B / 0 B
LAN tcp 127.0.0.1:8443 (10.10.10.1:443) <- 192.168.1.101:35770 CLOSED:SYN_SENT 2 / 0 120 B / 0 B
When loading 10.10.10.1 in a browser.

renegade

@BBcan177:

@pfcode:

There is a serious issue in the unbound service package, you will have to disable DNSBL, so that the line "server:include: /var/unbound/pfb_dnsbl.conf" is removed from Custom Options in the Unbound, to make unbound restart properly when your modem is rebooted or your IPv4/IPv6 WAN IP got renewed.

Please post this issue in Redmine. Its the best way to help get this resolved.

https://redmine.pfsense.org/projects/pfsense

Okay guys, what can i do here?
I am really sad that i cannot use DNSBL, because my provider cuts my line every day…
Do i have to post a bounty and refer to that forum post?

tjbaudio

I think I have traced my problem to NAT reflection. DNSBL is responding with the VIP address when tested with a known add site. The VIP server is up and responds as it should when using ports 8081 and 8443.
The firewall states show it forwarding when using 80 and 443. but nothing ever comes back.
I have tested this with wget, firefox, and chrom.

I love how DNSBL and pfBlockerNG are working. I am just going to have to go back over my pfsense setup to figure out the NAT reflection problem.

BBcan177

@renegade:

Okay guys, what can i do here?
I am really sad that i cannot use DNSBL, because my provider cuts my line every day…
Do i have to post a bounty and refer to that forum post?

Hi renegade,

You don't need to post a bounty. The redmine link is a site to post bugs/features for pfSense. The issue at hand seems to be related to the base pfSense software and not with the package. I don't have a similar network environment to debug this issue. So that's why I suggested that someone with the issue post in redmine to help sort out this issue. In the meantime try to reduce the number of feeds used for DNSBL which might reduce the unbound resolver load time.

BBcan177

@tjbaudio:

I think I have traced my problem to NAT reflection.

From an earlier post, it seems like you reversed the in/outbound interfaces in the pkg General tab. Not sure if that's related. Also try to enable suppression in the general tab and run a "Force Reload -All". This will remove any loopback addresses that could potentially be in the DNSBL IP aliastable.

tjbaudio

Thank you for the recommendations. These did not work. Unfortunately I may not have much time over the next week to work on this. I am thinking I may try a fresh install again when I do. Taking screen shots the whole way. If it works you can use them for a how to, if not I will know what I did :o

D0X

Problem here;

Running pfsense 2.3.1 with pfblockerng.

Did set up IPV4 with two lists, EasyList and EasyPrivacy.
I also have DNSBL enabled with Easylist.

But now when browsing I continuously get Safari telling me the certificate is invalid (as shown in the attached picture, it is in Dutch but I think you'll get it).
The certificate seems to me just the thing you'd like to have being blocked. How do I fix this, it is bugging the hell out of me.

![Schermafbeelding 2016-07-25 om 12.43.09.png](/public/imported_attachments/1/Schermafbeelding 2016-07-25 om 12.43.09.png)
![Schermafbeelding 2016-07-25 om 12.43.09.png_thumb](/public/imported_attachments/1/Schermafbeelding 2016-07-25 om 12.43.09.png_thumb)

BBcan177

@tjbaudio:

Thank you for the recommendations. These did not work. Unfortunately I may not have much time over the next week to work on this. I am thinking I may try a fresh install again when I do. Taking screen shots the whole way. If it works you can use them for a how to, if not I will know what I did :o

With just a single WAN/LAN, it should be fairly straightforward to setup the pkg… Double check your LAN Devices for DNS and ensure that they only have pfSense as its only DNS server setting.

BBcan177

@D0X:

But now when browsing I continuously get Safari telling me the certificate is invalid (as shown in the attached picture, it is in Dutch but I think you'll get it).

Safari seems to be the only browser that has this issue. Doesn't seem to occur with Chrome/FF/IE/Edge etc…

Are these cert popups happening when you browse to that blocked Domain directly, or when its blocking that Domain as part of the webpage? Two checks to ensure DNBSL is working as expected... Ping the DNSBL VIP and get a proper response ... Browse to the DNSBL VIP and get the 1x1 pix...

If the popups are from a few common Domains, try to whitelist those and see if that improves it.

Also, in the IPv4 tab, you cannot enter any EasyList Feeds... the IPv4 tab is for IP based listed only.

DNSBL is designed to block the DNS resolution to Advert/Malicious Domains. As such only a portion of the EasyList/Privacy Feeds can be used (Where the feeds lists the actual ADvert Domain). Browser-Addons like ADBlock/UBlock manipulate the elements that are displayed on the page. So they can block certain elements from a particular Domain, while DNSBL can only block the entire Domain.

Only two EasyList feeds can be used with DNSBL and they are hardcoded in the EasyList Tab. There are several other DNSBL Feeds listed at the start of this thread that can fill in the gap to block the balance of the Domains from serving any ADs. Plus its achieved at Network level, without any add-ons manipulating or potentially opening security holes in the browser.

It would really be nice if Safari fixed this issue with their software... Not much I can do to fix that without killing the package logging feature and just NXDOMAIN the DNS requests.

D0X

Safari seems to be the only browser that has this issue. Doesn't seem to occur with Chrome/FF/IE/Edge etc…

Are these cert popups happening when you browse to that blocked Domain directly, or when its blocking that Domain as part of the webpage? Two checks to ensure DNBSL is working as expected... Ping the DNSBL VIP and get a proper response ... Browse to the DNSBL VIP and get the 1x1 pix...

Should the VIP be one inside the LAN subnet or outside, tried both. Only inside the subnet I am getting the 1x1 pix, both don't respond to pings.

If the popups are from a few common Domains, try to whitelist those and see if that improves it.

googleads.g.doubleclick.net seems to be the main culprit. Isn't there any other way to block this certificate?

Also, in the IPv4 tab, you cannot enter any EasyList Feeds… the IPv4 tab is for IP based listed only.

Check, only using DNSBL now.

It would really be nice if Safari fixed this issue with their software… Not much I can do to fix that without killing the package logging feature and just NXDOMAIN the DNS requests.

BBcan177

@D0X:

Should the VIP be one inside the LAN subnet or outside, tried both. Only inside the subnet I am getting the 1x1 pix, both don't respond to pings.

Typically you set the DNSBL VIP to "LAN", and if you are in a Multi-LAN network, you can enable the option in the DNSBL Tab to create a Floating Permit rule to allow all of your LAN Subnets to hit the DNSBL VIP address. You will get unexpected behaviour until your fix the Ping/Browse to the VIP. Also ensure that your LAN devices only have pfSense as its DNS Server.

D0X

@BBcan177:

@D0X:

Should the VIP be one inside the LAN subnet or outside, tried both. Only inside the subnet I am getting the 1x1 pix, both don't respond to pings.

Typically you set the DNSBL VIP to "LAN", and if you are in a Multi-LAN network, you can enable the option in the DNSBL Tab to create a Floating Permit rule to allow all of your LAN Subnets to hit the DNSBL VIP address. You will get unexpected behaviour until your fix the Ping/Browse to the VIP. Also ensure that your LAN devices only have pfSense as its DNS Server.

Thanks for your support so far, but for now I've given up on pfblockerng. Not that the package is flawed, but Safari on OSX and IOS is driving me nuts.

BBcan177

@D0X:

Thanks for your support so far, but for now I've given up on pfblockerng. Not that the package is flawed, but Safari on OSX and IOS is driving me nuts.

NP…. Should you decide to pick it backup at some point, try to see why the Ping/Browse to the DNSBL VIP wasn't working. That will cause issues on its own.... also since it was one Domain that was predominantly causing issues, you could also Whitelist it...

Qinn

Hi there is it possible to save and restore the pfBlockerNG settings at a certain phase, so when you mess up you can return to a previous setup?

RonpfS

@Qinn:

Hi there is it possible to save and restore the pfBlockerNG settings at a certain phase, so when you mess up you can return to a previous setup?

Every change in pfBlockerNG is saved in the config file.

You can save the config with Diagnostics/Backup & Restore/Backup & Restore and do a restore of the config at a later time.

wiz561

I just updated the pfBlockerNG package and I'm now receiving these errors when I reboot…

There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:37:54
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:39:58
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:40:37

I'm running pfBlockerNG 2.1.1_2 with PfSense 2.3.2-RELEASE (amd64). I have 2gb of memory and 34% in use with a 2gb swap and 0% in use (for the swap).

I'm assuming it's not a memory issue and possibly something else. I was wondering if anybody else had this issue and what a possible fix may be.

Thanks!

RonpfS

@wiz561:

I just updated the pfBlockerNG package and I'm now receiving these errors when I reboot…
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:37:54
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:39:58
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:40:37
I'm running pfBlockerNG 2.1.1_2 with PfSense 2.3.2-RELEASE (amd64). I have 2gb of memory and 34% in use with a 2gb swap and 0% in use (for the swap).

I'm assuming it's not a memory issue and possibly something else. I was wondering if anybody else had this issue and what a possible fix may be.

Thanks!

Did you ran a Force Reload IP since the reboot?
Look at /var/db/aliastables/pfB_Europe_v6.txt, it should have only IP in it.

wiz561

@RonpfS:

@wiz561:
I just updated the pfBlockerNG package and I'm now receiving these errors when I reboot…
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:37:54
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:39:58
There were error(s) loading the rules: /tmp/rules.debug:47: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [47]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 12:40:37
I'm running pfBlockerNG 2.1.1_2 with PfSense 2.3.2-RELEASE (amd64). I have 2gb of memory and 34% in use with a 2gb swap and 0% in use (for the swap).

I'm assuming it's not a memory issue and possibly something else. I was wondering if anybody else had this issue and what a possible fix may be.

Thanks!
Did you ran a Force Reload IP since the reboot?
Look at /var/db/aliastables/pfB_Europe_v6.txt, it should have only IP in it.

Thanks for the response. Yes, I ran through the force reload (all of them) and (update, cron, reload), and nothing. I even tried to move the pfB_Europe_v6.txt file to .old and I'm still getting that error message.

RonpfS

https://forum.pfsense.org/index.php?topic=115357.msg643944#msg643944

@BBcan177:

@lpallard:
There were error(s) loading the rules: /tmp/rules.debug:37: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [37]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 14:55:00
There were error(s) loading the rules: /tmp/rules.debug:37: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [37]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 14:55:11
There were error(s) loading the rules: /tap/rules.debug:37: cannot define table pfB_Europe_v6: Cannot allocate memory - The line in question reads [37]: table <pfB_Europe_v6> persist file "/var/db/aliastables/pfB_Europe_v6.txt" @ 2016-07-31 14:55:20 
These all seem to be related to the MaxMind IPv6 database. Looks like you will need to bump the pfSense max aliastable entries limit from 2M to 4M. If you enable aggregation in the general tab, it should condense the CIDRs and reduce the overall IP count. This changed due to using the new MaxMind Geolite2 database which seems to have smaller subsets of the data listed causing more IP entries to be added.