Unofficial E2guardian package for pfSense

marcelloc

Try to disable watch guard and see on access log which was the last site or url accessed before crash. Repeat few times to see if we can get close to a way to reproduce the crash on any system.

marcelloc

to install a debug version of e2guardian, paste this under console

pkg add -f https://e-sac.websiteseguro.com/e2g/e2guardian-5.0.2_5.txz

then run

/usr/local/sbin/e2guardian -QN

to reinstall current Unofficial repo version use

pkg install -f e2guardian

pfsensation

@marcelloc:

Try to disable watch guard and see on access log which was the last site or url accessed before crash. Repeat few times to see if we can get close to a way to reproduce the crash on any system.

Great idea, I'll definitely give that a go now. Thanks for your input!

pfsensation

@marcelloc:

to install a debug version of e2guardian, paste this under console
pkg add -f https://e-sac.websiteseguro.com/e2g/e2guardian-5.0.2_3.txz
then run

/usr/local/sbin/e2guardian -QN

to reinstall current Unofficial repo version use
pkg install -f e2guardian

Can we not integrate this into the unofficial rep? And allow the debug mode to be turned on or off? I think that maybe easier to manage and use.

I've tried a lot of things last night to try combat the crashes, I've even changed my pfBlocker ports again even though I've never set them up to conflict with each other. I did see in logs that E2 Guardian was saying that port 8080 was in use. I'm not sure if it was just ghost instances of itself or another service. Nothing else is set to use my E2G ports.

marcelloc

@pfsensation:

Can we not integrate this into the unofficial rep? And allow the debug mode to be turned on or off? I think that maybe easier to manage and use.

No. It's a compile option.

marcelloc

@pfsensation:

I did see in logs that E2 Guardian was saying that port 8080 was in use. I'm not sure if it was just ghost instances of itself or another service. Nothing else is set to use my E2G ports.

Probably yes. Parallel e2guardian proceses due watch guard script trying to start e2guardian during an apply config or another apply config while process is reloading…

pfsensation

So I've just had E2Guardian crash on me again… This time I have snapped all the logs. Here's the last few lines on the access.log

18.04.15 16:19:11 172.16.1.6 172.16.1.6 https://r1---sn-8pgbpohxqp5-ajtl.googlevideo.com:443 *TRUSTED* Site match: googlevideo.com CONNECT 138829 0 - 3 200 - 172.16.1.6 wireless_users - - - - -
18.04.15 16:19:13 172.16.1.3 172.16.1.3 https://use.fontawesome.com - - 5060 0 - 2 0 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:15 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=01c53a8c-ce4e-4a87-948a-bb420093349e - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:18 172.16.1.6 172.16.1.6 https://m.youtube.com/api/stats/watchtime?ns=yt&el=detailpage&cpn=Bnm_BnpPYGGmvaw6&docid=9YZgy1dpesY&ver=2&referrer=https%3A%2F%2Fm.youtube.com%2Fwatch%3Fv%3DQFe1PcLZGQE%26itct%3DCBYQpDAYAiITCPGnsKHIvNoCFRekFQodY%252BkPo1IKbWFkcyBsZXdpcw%253D%253D%26lact%3D12&cmt=155.448&plid=AAVp5J4QQ2VIWDNA&ei=RG3TWvv3LdbzWtWSgZgM&fmt=134&fs=0&rt=49.006&of=sidReqaeP4yRcfYsI6hTkw&euri&lact=39362&cl=192651238&state=playing&vm=CAEQABgE&volume=100&c=MWEB&cver=1.20180412&cplayer=UNIPLAYER&cbrand=generic&cbr=Firefox&cbrver=59.0&cmodel=android%204.4&cos=Android&cosver=4.4.4&cplatform=TABLET&hl=en_GB&cr=GB&len=796.261&rtn=89&feature=related&afmt=251&idpj=-4&ldpj=-25&rti=49&muted=0&st=136.465&et=155.448&conn=3 - GET 0 0 - 3 204 - 172.16.1.6 wireless_users - - - - -
18.04.15 16:19:20 172.16.1.6 172.16.1.6 https://m.youtube.com/api/stats/qoe?event=streamingstats&fmt=134&afmt=251&cpn=Bnm_BnpPYGGmvaw6&ei=RG3TWvv3LdbzWtWSgZgM&el=detailpage&docid=9YZgy1dpesY&ns=yt&fexp=23707874%2C23708904%2C23708906%2C23708910%2C23710476%2C23712544%2C23712994%2C23713711%2C23716256%2C23716972%2C23718632%2C23720634%2C23721898%2C23723618%2C23723926%2C23724493%2C23725949%2C23726426%2C23728352%2C23731308%2C23731896%2C23732330%2C23732385%2C23732729%2C23733751%2C23734535%2C23735135%2C23735347%2C23736055%2C23736249%2C23736874%2C9422596%2C9449243%2C9475691%2C9477414%2C9485000%2C9486081%2C9487194%2C9489759&cl=192651238&seq=4&c=MWEB&cver=1.20180412&cplayer=UNIPLAYER&cbrand=generic&cbr=Firefox&cbrver=59.0&cmodel=android%204.4&cos=Android&cosver=4.4.4&cplatform=TABLET&vps=50.154:PL&bwm=50.154:3569482:1.432&bwe=50.154:1955673&cmt=50.154:156.525&bh=50.154:51.736 - POST 0 0 - 3 204 - 172.16.1.6 wireless_users - - - - -
18.04.15 16:19:21 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=6a8e596b-71f5-4665-91c3-b18339d68777 - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:21 172.16.1.6 172.16.1.6 https://endpoint1.collection.eu.sumologic.com/receiver/v1/http/ZaVnC4dhaV3f1zhS5xhQcAjcICG2O_Vtd8mu-39iqkdF1iAhhFWHpSA5bUBK_SrbgvG8GdcqjH_DGs-NplOisVzYbr-qoEfcff0MhUJH4RV79DiuIFEF0w== - POST 0 0 - 3 200 - 172.16.1.6 wireless_users - - - - -
18.04.15 16:19:22 172.16.1.3 172.16.1.3 https://easylist.to - - 148068 0 - 2 0 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:25 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=01c53a8c-ce4e-4a87-948a-bb420093349e - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:29 172.16.1.7 172.16.1.7 https://play.googleapis.com *TRUSTED* Site match: play.googleapis.com - 1166 0 - 4 0 - 172.16.1.7 wireless_users - - - - -
18.04.15 16:19:29 172.16.1.6 172.16.1.6 https://s3-eu-west-1.amazonaws.com/smhw-frontend-prod/VERSION.json - GET 0 0 - 3 304 - 172.16.1.6 wireless_users - - - - -
18.04.15 16:19:31 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=6a8e596b-71f5-4665-91c3-b18339d68777 - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:34 172.16.1.3 172.16.1.3 http://go.microsoft.com/fwlink/?LinkID=252669&clcid=0x409 - POST 0 0 - 2 302 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:35 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=01c53a8c-ce4e-4a87-948a-bb420093349e - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:36 172.16.1.6 172.16.1.6 https://m.youtube.com/youtubei/v1/log_event?alt=json&key=8Q4STEHLGCilw_Y9_11qcW8 - POST 48 0 - 3 200 - 172.16.1.6 wireless_users - - - - -
18.04.15 16:19:41 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=6a8e596b-71f5-4665-91c3-b18339d68777 - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:45 172.16.1.3 172.16.1.3 http://push.bitdefender.net/poll?push_id=01c53a8c-ce4e-4a87-948a-bb420093349e - GET 16 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:45 172.16.1.3 172.16.1.3 https://nexus.officeapps.live.com:443 - CONNECT 152 0 - 2 200 - 172.16.1.3 Forid - - - - -
18.04.15 16:19:46 172.16.1.3 172.16.1.3 https://beacons5.gvt3.com - - 3362 0 - 2 0 - 172.16.1.3 Forid - - - - -

We've got a tasty collection of enormous URL's, post requests and a tonne of random telemetry. @Marcelloc does this help?

marcelloc

Can you run e2g with debug binaries, browse these latest sites to get the error message?

what authentication are you using to get you user on logs?

pfsensation

@marcelloc:

Can you run e2g with debug binaries, browse these latest sites to get the error message?

what authentication are you using to get you user on logs?

I just had another crash, see below:

18.04.15 19:08:18 172.16.1.3 172.16.1.3 https://client.wns.windows.com - - 4430 0 - 2 0 - 172.16.1.3 Forid - - - - -
18.04.15 19:09:05 172.16.1.17 172.16.1.17 http://www.msftconnecttest.com/connecttest.txt - GET 22 -40 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:12 172.16.1.17 172.16.1.17 http://cdn.content.prod.cms.msn.com/singletile/summary/alias/experiencebyname/today?market=en-GB&source=appxmanifest&tenant=amp&vertical=finance - GET 1008 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:12 172.16.1.17 172.16.1.17 http://cdn.content.prod.cms.msn.com/singletile/summary/alias/experiencebyname/today?market=en-GB&source=appxmanifest&tenant=amp&vertical=sports - GET 1137 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:13 172.16.1.17 172.16.1.17 http://cdn.content.prod.cms.msn.com/singletile/summary/alias/experiencebyname/today?market=en-GB&source=appxmanifest&tenant=amp&vertical=news - GET 1248 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:16 172.16.1.17 172.16.1.17 http://cdn.content.prod.cms.msn.com/singletile/summary/alias/experiencebyname/today?market=en-GB&source=appxmanifest&tenant=amp&vertical=sports - GET 1137 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:16 172.16.1.17 172.16.1.17 http://cdn.content.prod.cms.msn.com/singletile/summary/alias/experiencebyname/today?market=en-GB&source=appxmanifest&tenant=amp&vertical=finance - GET 1008 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:16 172.16.1.17 172.16.1.17 http://cdn.content.prod.cms.msn.com/singletile/summary/alias/experiencebyname/today?market=en-GB&source=appxmanifest&tenant=amp&vertical=news - GET 1248 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:21 172.16.1.17 172.16.1.17 https://client-office365-tas.msedge.net - - 12400 0 - 2 0 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:21 172.16.1.17 172.16.1.17 https://config.edge.skype.com *TRUSTED* Site match: skype.com - 6894 0 - 2 0 - 172.16.1.17 Forid - - - - -
18.04.15 19:09:41 - 172.16.1.10 https://mmg-fna.whatsapp.net:443 *TRUSTED* Site match: whatsapp.net CONNECT 358 0 - 1 200 - 172.16.1.10 Default - - - - -
18.04.15 19:09:58 172.16.1.17 172.16.1.17 https://client.wns.windows.com - - 4430 0 - 2 0 - 172.16.1.17 Forid - - - - -
18.04.15 19:10:05 - 172.16.1.14 http://clients3.google.com/generate_204 *TRUSTED* Site match: clients3.google.com GET 0 0 - 1 204 - 172.16.1.14 Default - - - - -
18.04.15 19:10:05 - 172.16.1.14 http://clients3.google.com/generate_204 *TRUSTED* Site match: clients3.google.com GET 0 0 - 1 204 - 172.16.1.14 Default - - - - -
18.04.15 19:10:07 172.16.1.17 172.16.1.17 https://nexus.officeapps.live.com:443 - CONNECT 2904 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:10:07 172.16.1.17 172.16.1.17 https://nexusrules.officeapps.live.com:443 - CONNECT 2904 0 - 2 200 - 172.16.1.17 Forid - - - - -
18.04.15 19:10:07 - 172.16.1.14 https://api.samsungcloud.com:443 - CONNECT 6400 0 - 1 200 - 172.16.1.14 Default - - - - -
18.04.15 19:10:08 - 172.16.1.14 https://api.samsungcloud.com:443 - CONNECT 383 0 - 1 200 - 172.16.1.14 Default - - - - -
18.04.15 19:10:12 172.16.1.17 172.16.1.17 https://v20.vortex-win.data.microsoft.com - - 4215 0 - 2 0 - 172.16.1.17 Forid - - - - -

I am using IP authentication, with groups. I've got static IP for anyone who needs any "special" ACL's added. I will install the debug binaries now.

pfsensation

@Maracello, I have installed the debug binaries and it's made the mystery deeper! I have tried all the last sites in the access.log and had no crashes… I would set the debug logs to be written to a file but I can tell that they would become huge in a matter of minutes.

Have you got any other suggestions?

marcelloc

Leave it open without redirect on a ssh console.

When it crashes, latest info on ssh console will be the relevant information to send to e2guardian dev team.

pfsensation

@marcelloc:

Leave it open without redirect on a ssh console.

When it crashes, latest info on ssh console will be the relevant information to send to e2guardian dev team.

Got it, thanks bro! I'll keep a SSH session open with this on my ESXi VM :)

EDIT:

GOTCHA!! Found the problem, it's the HOSTS in the block page. Removing that broken feature right away.

hw2390: Start of request header:in
hw2390: header:in before getLine - timeout:55000 Line: 1938 Function: in
hw2390: firstime: header:in after getLine  Line: 1942 Function: in
hw2390: header:size =  7
hw2390: first line =  POST /vpninfo/servers HTTP/1.1
 Line: 981 Function: checkheader: Host: www.privateinternetaccess.com
 Line: 981 Function: checkheader: Accept-Encoding: identity,gzip,deflate
 Line: 981 Function: checkheader: Accept: */*
 Line: 981 Function: checkheader: User-Agent: Ruby
hw2390: tp =Content-Length: 49
hw2390: tp =49 Contentlen.int =49
 Line: 981 Function: checkheader: Content-Length: 49
hw2390: Header value from client: Content-Type: application/x-www-form-urlencode Line: 981 Function: checkheader
hw2390: CheckHeader: HTTP/1.1 detected Line: 990 Function: checkheader
hw2390: CheckHeader flags before normalisation: AP=1 PPC=0 1.1=1 connectionclose=1 CL=1 Line: 1051 Function: checkheader
hw2390: CheckHeader flags after normalisation: AP=0 WP=0 Line: 1074 Function: checkheader
hw2390: CheckHeader: Adding our own Proxy-Connection: Close Line: 1093 Function: checkheader
hw2390: HTTPHeader size: 8
hw2390: from header url:http://www.privateinternetaccess.com/vpninfo/servers Line: 1235 Function: getUrl
hw2390: setURL: header.front() changed from: POST http://www.privateinternetacce Line: 540 Function: setURL/1.1
 Line: 548 Function: setURLw.privateinternetaccess.com/vpninfo/servers HTTP/1.1
 Line: 553 Function: setURLne changed from: Host: www.privateinternetaccess.com
 Line: 562 Function: setURLteinternetaccess.com
hw2390: isProxyRequest is 0
hw2390:  firsttime =1ismitm =0 clientuser = group = 1
hw2390: Compiling ,*[a-z|A-Z].*
hw2390: ...with PCRE
hw2390: HTTPHeader size: 8
hw2390: from header url:http://www.privateinternetaccess.com/vpninfo/servers Line: 1235 Function: getUrl
hw2390: decoding url Line: 1519 Function: decode
hw2390: 38420Start URL http://www.privateinternetaccess.com/vpninfo/serversis_ssl=0ismitm=0
hw2390: inIPList no match for 172.16.1.17
hw2390: inIPList no match for 172.16.1.17
hw2390: inURLList: www.privateinternetaccess.com/vpninfo/servers
hw2390: inURLList (processed): www.privateinternetaccess.com/vpninfo/servers
After StoryA pre-authcheck0 mess_no 0
hw2390: isProxyRequest is 0 only_ip_auth is 1
hw2390:  -Not got persistent credentials for this connection - querying auth plugins
hw2390:  -Querying next auth plugin...
hw2390:  -Auth plugin found username 172.16.1.17 (), now determining group
hw2390: Matched IP 172.16.1.17 to straight IP list
hw2390: Auth plugin found username & group; not querying remaining plugins
hw2390:  -Identity found; caching username & group
Auth plugin is for a connection-based auth method - keeping credentials for entire connection
hw2390:  -username: 172.16.1.17
hw2390:  -filtergroup: 1
hw2390:  -About to check for bypass...
hw2390: Found cookie: Line: 1309 Function: getCookie
hw2390: No bypass cookie Line: 1337 Function: isBypassCookie
hw2390:  -Finished bypass checks.
hw2390: inURLList: www.privateinternetaccess.com/vpninfo/servers
hw2390: inURLList (processed): www.privateinternetaccess.com/vpninfo/servers
hw2390: inURLList: www.privateinternetaccess.com/vpninfo/servers
hw2390: inURLList (processed): www.privateinternetaccess.com/vpninfo/servers
hw2390: After StoryB checkrequest isexception 0 gomitm 0 mess_no 500
hw2390:  -Forwarding body to client : Upfailure is 0
hw2390:  -reporting level is 3
hw2390:  -Enabling filter bypass hash generation
hw2390: mime type: - Line: 217 Function: isContentType
hw2390: mimes result : false Line: 244 Function: isContentType
hw2390: Displaying TEMPLATE
hw2390: -HOST- placeholder encountered but hostname currently unknown; lookup forced.
Segmentation fault

marcelloc

Happened and fixed on 4.1. maybe not merged on 5.0

I'll se if I find the issue number on 4.1 branch and post on #375

pfsensation

@marcelloc:

Happened and fixed on 4.1. maybe not merged on 5.0

I'll se if I find the issue number on 4.1 branch and post on #375

Face palm even without the -HOSTS- place holder in my block page. E2 Guardian still crashes after sometime, I guess I'll have to do more debugging tomorrow.

pfsensation

@pfsensation:

@marcelloc:

Happened and fixed on 4.1. maybe not merged on 5.0

I'll se if I find the issue number on 4.1 branch and post on #375

Face palm even without the -HOSTS- place holder in my block page. E2 Guardian still crashes after sometime, I guess I'll have to do more debugging tomorrow.


SB  list name banned checking type  3
Looking for banned type 3 in listmeta
Found banned type 3 in listmeta
list reference 196 'banned' found for banned
SB  list name banned checking type  2
Looking for banned type 2 in listmeta
Found banned type 2 in listmeta
list reference 206 'banned' found for banned
SB  list name banned checking type  6
Looking for banned type 6 in listmeta
Found banned type 6 in listmeta
list reference 223 'banned' found for banned
banned matches 3 types
Line 218 state is 2 actionid 5003 listname banned
Compiling .*[a-z|A-Z].*
...with PCRE
Setting magic key to 'PWYDSKLILPOTMCAK'
read filter group: 4 /usr/local/etc/e2guardian/e2guardianf5.conf return is 1
I seem to be running already!
In deleteFilterGroups loop
In deleteFilterGroups loop
In deleteFilterGroups loop
In deleteFilterGroups loop
In deleteFilterGroups loop

Just got this @Marcelloc and then E2G crashed. Weird… Not sure what return is 1 means.

Also getting the below with E2G Debug binaries:

Apr 16 14:49:22	e2guardian	83603	Tunnel status not 200 ok aborting
Apr 16 14:49:24	e2guardian	83603	Tunnel status not 200 ok aborting
Apr 16 14:49:30	e2guardian	83603	Tunnel status not 200 ok aborting
Apr 16 14:49:33	e2guardian	83603	Tunnel status not 200 ok aborting

marcelloc

I seem to be running already! Is the error when there is another process running (apply whike reloading, watchdog, etc…)

pfsensation

@marcelloc:

I seem to be running already! Is the error when there is another process running (apply whike reloading, watchdog, etc…)

I think the watchdog may need to be re-designed, because that shouldn't happen. My E2G is set to kill a running instances and start new ones on config change.

pfsensation

Hi Marcelloc,

Could you please update the binaries, the E2G developers have pushed out some patches which should help with these segmentation faults. Funnily enough the debug version seems to work more stable for me than the one in the repo.

Thanks

landforces

hi marcello

There are 2 responsibilities

1. exe zip vs download blocking

When we try to enter some prohibited sites, such as the second picture, it reacts differently
one says that this site can not be reached and the other page comes Block.

e2guard.PNG_thumb

marcelloc

@pfsensation:

Hi Marcelloc,

Could you please update the binaries, the E2G developers have pushed out some patches which should help with these segmentation faults. Funnily enough the debug version seems to work more stable for me than the one in the repo.

Thanks

Binaries updated on repo.