Guest User!

You are not Sophos Staff.

Thread Info
  • State Verified Answer
  • +2 person also asked this people also asked this
  • Locked Locked
  • Replies 14 replies
  • Answers 2 answers
  • Subscribers 5 subscribers
  • Views 63645 views
  • Users 0 members are here
Options
Suggested
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Remote access SSL VPN not working

MikeJeffers

Hi folks. I am trying to establish a Remote Access VPN tunnel on my newly deployed UTM 9 Home. The configuration for the SSL VPN is pretty straightforward and I've followed all of the instructions that Sophos published here.

 

The remote browser gives a different error message depending on the browser used, but I found this in the Live Log while trying to connect. I've sanitized the output. 1.1.1.1 is the remote client browser and 2.2.2.2 is the public IP of the UTM.

I'm not sure what the MTU would have to do with anything. Strangely the MTU on my external interface was 576. I upped it to 1475.

 

10:23-22:20:59 home-utm openvpn[1574]: TCP connection established with [AF_INET]1.1.1.1:2652 (via [AF_INET]2.2.2.2:443)

2016:10:23-22:20:59 home-utm openvpn[1574]: 1.1.1.1:2652 WARNING: Bad encapsulated packet length from peer (5635), which must be > 0 and <= 1563 -- please ensure that --tun-mtu or --link-mtu is equal on both peers -- this condition could also indicate a possible active attack on the TCP link -- [Attempting restart...]

2016:10:23-22:20:59 home-utm openvpn[1574]: 1.1.1.1:2652 Connection reset, restarting [0]

2016:10:23-22:20:59 home-utm openvpn[1574]: 1.1.1.1:2652 SIGUSR1[soft,connection-reset] received, client-instance restarting

2016:10:23-22:20:59 home-utm openvpn[1574]: TCP connection established with [AF_INET]1.1.1.1:2317 (via [AF_INET]2.2.2.2:443)

2016:10:23-22:20:59 home-utm openvpn[1574]: 1.1.1.1:2317 WARNING: Bad encapsulated packet length from peer (5635), which must be > 0 and <= 1563 -- please ensure that --tun-mtu or --link-mtu is equal on both peers -- this condition could also indicate a possible active attack on the TCP link -- [Attempting restart...]

2016:10:23-22:20:59 home-utm openvpn[1574]: 1.1.1.1:2317 Connection reset, restarting [0]

2016:10:23-22:20:59 home-utm openvpn[1574]: 1.1.1.1:2317 SIGUSR1[soft,connection-reset] received, client-instance restarting



This thread was automatically locked due to age.
Parents
  • +1

    Hi Mike,

    maybe you run into a bug with mtu size of the external interface. do you use a cable provider?

    check your mtu of external again. think you cant change it with the webinterface..

    in this post there is fix described

    https://community.sophos.com/products/unified-threat-management/f/hardware-installation-up2date-licensing/80641/sophos-utm-9-407-3-released

    check that first and check your external mtu then to get a stable connection. after that can check ssl-vpn

    greets

    zaphod
    ___________________________________________

    Home: Zotac CI321 (8GB RAM / 120GB SSD)  with latest Sophos UTM
    Work: 2 SG430 Cluster / many other models like SG105/SG115/SG135/SG135w/...

  • 0 in reply to zaphod

    I'm such a ding dong.

     

    I was able to leverage that thread and disable the MTU auto-discover allowing me to manually set it to 1500 from 576. So that solved many other problems, I'm sure.

     

    The error message continued in the LiveLog:

    WARNING: Bad encapsulated packet length from peer (5635), which must be > 0 and <= 1563 -- please ensure that --tun-mtu or --link-mtu is equal on both peers -- this condition could also indicate a possible active attack on the TCP link -- [Attempting restart...]

     

    The problem was completely unrelated to MTU. I failed to enable the User Portal. I enabled it and boom.... the SSL VPN answered and I was able to remotely connect. <face palm>

     

     

     

  • 0 in reply to MikeJeffers

    Thanks for the lead on the MTU bug. That was HUGE. I was having issues with Outlook 2016 client sending email via RPC over HTTP to my Hosted Exchange deployment with Office365. That's now working.

    General web browsing performance was really bad. Pages are loading much faster now.

    Online FPS gaming on my PS4 was horribly bad. Unplayable actually. That's working fine now.

     

  • 0 in reply to MikeJeffers

    Hi Mike,

     

    glad to help you fixing this out. the solution need to be pinned its difficult to find.

     

    please mark my suggestion as answer for you so that others can see thread is answered. thanks.

    greets

    zaphod
    ___________________________________________

    Home: Zotac CI321 (8GB RAM / 120GB SSD)  with latest Sophos UTM
    Work: 2 SG430 Cluster / many other models like SG105/SG115/SG135/SG135w/...

  • 0 in reply to zaphod

    Again, many thanks for the lead on the MTU bug.

     

    Re: The Answer

     

    The answer to the original problem (Remote Access SSL VPN not working) was not related to the MTU size despite the Live Log pointing us in that direction. The SSL VPN was not working, because I had not enabled the User Portal. Once I enabled the User Portal the SSL VPN would then answer the HTTPS client request.

  • 0 in reply to MikeJeffers

    Hi Mike Jeffers,

    normally you dont have to activate the User Portal to use the SSL-VPN.

    Best regards

    DKNL

Reply Children
  • 0 in reply to DKKDG

    Hello,

     

    I have more or less same issue with UTM SSL VPN. I can though connect to VPN but unable to access the remote instance/machines after connecting. I have UTM appliance configured on AWS on external public subnet and want to connect to my internal instances. Can anyone suggest something. I have this kind of setup already on my other VPCs and working fine but this one seems to have some issue or i have some misconfiguration.

     

    Thanks, 

  • 0 in reply to techieh

    Does doing #1 in Rulz provide any insight?

    Cheers - Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA
  • 0 in reply to BAlfson

    Hi Bob,

     

    I have already disabled IPS (checked the tabs as well). Intrusion Prevention , Application Control and Firewall logs does not show anything either.  I 've attached all the screenshots possible. 

     

    1.  

    2. 

     

    3. 

     

    4.

     

     

    5

    . 

     

    6. 

  • +1 in reply to techieh

    Thanks, Vibhor, that really madee it easy to help you.

    Since there was no indication in the logs, I anticipated finding a routing problem.  I suspect that adding "LR Subnet" to 'Local Networks' in the SSL VPN Profile will cause WebAdmin to create the necessary routes.  This will also make firewall rule #4 redundant.

    I don't think the SNAT has any effect.

    The "Internal Networks" group must include the "VPN Pool (SSL)" object or another masq rule is needed for it.

    Firewall rule #2 is redundant if the following conditions are met:

    • DNS Forwarders are set correctly
    • The "VPN Pool (SSL) and "LR Subnet" objects are in DNS 'Allowed Networks'
    • 'Remote Access >> Advanced' is configured correctly

    Cheers - Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA
  • 0 in reply to BAlfson

    Hi Bob,

     

    Thanks for the info , i tried everything you mentioned 

    1. Added LR Subnet to Local networks in SSL VPN Profile 

    2. Turned Off rule #4 in firewall

    3. Added VPN Pool (SSL) in Internal Networks Object definition

    4 . DNS Forwarders are set correctly

    5.  The "VPN Pool (SSL) and "LR Subnet" objects are in DNS 'Allowed Networks'

    6.  Remote Access >> Advanced' is configured correctly

    After doing above disconnected VPN and tried again after reconnecting but its still the same.  Just for information DNS resolution is working fine after i connect to VPN. 

    Just the instances are not reachable  and that surely seems to be a routing issue but where i 'm not able to figure out. Need your expert opinion.

     

    Attached are the various screenshots

     

     

     

     

     

     

  • 0 in reply to techieh
    1. If you also want the SSL VPN users to access the Internet via the Tunnel, then you must add the "Internet" object.
    2. Does the pool of IPs for the SSL VPN, 10.242.2.0/24, conflict with any other subnets defined in your environment?
    3. I would have expected the first IP on RA_Advanced to be the IP of "Internal (Address)."  The second would be that of your AWS DNS which appears to be what you now have in the first position.

    Once these things are done, the UTM is correctly configured.  You could go over the routing table with a fine-tooth comb to confirm there are no conflicts, but I believe that any further changes to fix routing and firewall issues will need to be in your VPC.  At least, that's where I'd spend time proving that the issue isn't some other mis-configuration in the UTM.

    You should go ahead and get a case started with Sophos Support so that they can put someone inside your UTM instance to look around for other issues.

    Cheers - Bob

     
    Sophos UTM Community Moderator
    Sophos Certified Architect - UTM
    Sophos Certified Engineer - XG
    Gold Solution Partner since 2005
    MediaSoft, Inc. USA
Unfiltered HTML