Thread Info
  • State Verified Answer
  • Locked Locked
  • Replies 3 replies
  • Subscribers 5 subscribers
  • Views 9681 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

Slow Upload from Hyper-V instance of UTM 9.405-5

BrianGleason

I am seeing a similar thing as in this post:
community.sophos.com/.../73298

I have a 250M/25M connection and when I hook up a laptop directly to the internet connection, I get full BW up and down. When I hook the inet connection into the Sophos UTM, I am only getting approx 250M/3M. That upload rate coming from the UTM is not reasonable, and I mean that is a hard 3M, speed tests trickle up to that speed and just appear to stall like there is a throttle in place...

I am running Sophos UTM 9.405-5 on Hyper-V, I have a dedicated (not shared with the host) nic for the connection to the internet. I am not able to modify speed/duplex of the virtual nic in sophos and I have set the physical interface to 1G/Full with still no luck... It is something in the interaction with UTM and the connection. No, QoS is not turned on.

There has to be some resolution for this...

I have even tried the to drop a switch between the inet connection and the sophos interface and set both of those ports to 1G/Full just in case there was something there as I have read in other articles, only thing is they were not running on Hyper-V. I am not running any other services other than Firewall...

Thanks for any help in advanced..



This thread was automatically locked due to age.
  • 0

    I have validated that the issue lies within the SOPHOS UTM VM, but the question is what...?

    Since my UTM is on a HyperV host, I thought that maybe it has to do with the vNic, vSwitch, duplexing or something. I have another guest on that host that I have attached to the vNic/vSwitch, and on that VM I get full upstream/downstream like I should. So there is something in the UTM that is preventing me from getting full upsteam... I dont know if it is a residual configuration that is hanging around or what... Any ideas of where to look would be awesome...

    Down/Up
    Physical laptop: 262.65M/29.35M
    Guest VM(HyperV: Same Host/vswitch as UTM): 258.33M/29.97M
    Sophos UTM(HyperV): 278.68M/3.75M

    Thanks,

    Brian

  • 0 in reply to BrianGleason

    Additional Notes:

    I have create a 2012 R2 RRAS server and put that between the UTM and the ISP via a LAN/WAN NAT. When testing directly from the RRAS Server, I get the full 250M/25M (down/up), but from a device behind the UTM I get 250M/10-14M, again I expect some loss with the additional hops and translation.

    So far:

    Physical(Laptop) -> Modem -> Inet: 250M/25M
    HyperV Guest(Windows) -> Modem -> Inet: 250M/25M
    Client -> HyperV UTM -> HyperV Guest(Window NAT) -> Inet: 250M/10-14M
    Client -> HyperV UTM -> Modem -> Inet: 250M/3-5M

    I doubt anyone from Sophos would pick this up, but something is not right and I have no other blocking features on but FW... Maybe someone will come back with something.

    Thanks...

  • 0 in reply to BrianGleason

    The solution for this issue has been found... a bug introduced by a new forced configuration...:
    https://community.sophos.com/products/unified-threat-management/f/52/p/79288/304802#304802

    Making this change solved it for me...

    Just in case you need help with editing the file, here is some info on "vi", linux text editor..
    http://www.wikihow.com/Create-and-Edit-Text-File-in-Linux-by-Using-Terminal

    Thanks,

    Brian

Unfiltered HTML