Page 10 of 12 FirstFirst ... 89101112 LastLast
Results 91 to 100 of 119
  1. #91
    Untangler
    Join Date
    Jan 2020
    Location
    San Jose, CA
    Posts
    89

    Default

    Quote Originally Posted by theoak View Post
    Now there are a lot of factors that can contribute ...

    My buffer bloat scores have been amazing with this beta on my 50/3 (Mbps Down/Up) bonded pair DSL.

    I typically score all over the place any where from an A to a C with the bulk in the B to C score range. Typically the upload test would kill me from a buffer bloat score perspective.

    Since moving to the beta ... I have not gotten less than an A. It seems much better at my slower speeds.
    What tool are you using to measure buffer bloat? And - in case it doesn't run under Windows 10 - do you happen to know a tool that does?

  2. #92
    Master Untangler
    Join Date
    Aug 2016
    Posts
    230

    Default

    Quote Originally Posted by tangofan View Post
    What tool are you using to measure buffer bloat? And - in case it doesn't run under Windows 10 - do you happen to know a tool that does?
    dslreports web site speed test.
    Jim.Alles likes this.
    Untangle 16.0.1 (Build: 20201011T124207) (Kernel: 4.19.0-8-untangle-amd64)
    QOTOM-Q355G4
    1.6-2.7 GHz Intel I5 5250U, 128GB SSD mSATA, 8GB RAM DDR3L, 4xRJ-45 Intel I211AT 10/100/1000 Controller

  3. #93
    Untangle Ninja
    Join Date
    May 2008
    Posts
    1,290

    Default

    I believe he used dslreports.com speed test. Unfortunately you now need to be a member and have a history. LOL

  4. #94
    Untangle Ninja Jim.Alles's Avatar
    Join Date
    Jul 2008
    Location
    Central PA
    Posts
    2,523

    Default

    Flent is Linux
    A blog article: https://blog.tohojo.dk/2017/04/the-s...rk-tester.html

    An article about DSLreports: https://www.internetsociety.org/blog...om-dslreports/

    You can do your own comparative feel 'quick test' using NGFW and a client machine, as described here: https://www.bufferbloat.net/projects/bloat/wiki/Tests_for_Bufferbloat/

    On a windows client command line, start pinging Google:
    Code:
    ping -t 8.8.8.8
    The 'time=' reported number is latency.

    then start any speed test
    Give the buffers time to fill, then watch the variation in ping latency time during the upload phase.

    To compare, turn fq_codel on and off in NGFW #config/network/advanced/qos [Queue Discipline]
    You can also downward-adjust your WAN upload bandwidth, to avoid filling upstream buffers.

    This would be good to see in a graph, therefor Flent
    Last edited by Jim.Alles; 05-11-2020 at 12:55 PM.
    donhwyo likes this.

  5. #95
    Untangler
    Join Date
    Jan 2020
    Location
    San Jose, CA
    Posts
    89

    Default

    Quote Originally Posted by donhwyo View Post
    I believe he used dslreports.com speed test. Unfortunately you now need to be a member and have a history. LOL
    I was able to run the test and get my buffer bloat score without being a member.
    Jim.Alles likes this.

  6. #96
    Untangler
    Join Date
    Jan 2020
    Location
    San Jose, CA
    Posts
    89

    Default

    Quote Originally Posted by theoak View Post
    dslreports web site speed test.
    Thanks. I actually managed to get an "A" score, but had to lower my QoS upload speed to only 40% of my nominal speed. I guess Comcast is having difficulties providing enough upload capacity during daytime hours.

  7. #97
    Untangler
    Join Date
    Jan 2020
    Location
    San Jose, CA
    Posts
    89

    Default

    Quote Originally Posted by Jim.Alles View Post
    Flent is Linux
    A blog article: https://blog.tohojo.dk/2017/04/the-s...rk-tester.html

    An article about DSLreports: https://www.internetsociety.org/blog...om-dslreports/

    You can do your own comparative feel 'quick test' using NGFW and a client machine, as described here: https://www.bufferbloat.net/projects/bloat/wiki/Tests_for_Bufferbloat/

    On a windows client command line, start pinging Google:
    Code:
    ping -t 8.8.8.8
    The 'time=' reported number is latency.

    then start any speed test
    Give the buffers time to fill, then watch the variation in ping latency time during the upload phase.

    To compare, turn fq_codel on and off in NGFW #config/network/advanced/qos [Queue Discipline]
    You can also downward-adjust your WAN upload bandwidth, to avoid filling upstream buffers.

    This would be good to see in a graph, therefor Flent
    Cool, I will give this a try.

    Not long ago I've had serious problems with my GF trying to use Cisco AnyConnect VPN to connect to her work place and only after adding a bypass rule for her computer, did it finally work without multi-second delays, when typing in a bash shell in VNC Viewer. But as soon as I start any upload tasks, the delay comes back, no matter how low I set the upload bandwidth in QoS.

    Would be great, if this were indeed to get better in Untangle 15.1.

  8. #98
    Master Untangler
    Join Date
    Aug 2016
    Posts
    230

    Default

    Quote Originally Posted by tangofan View Post
    Thanks. I actually managed to get an "A" score, but had to lower my QoS upload speed to only 40% of my nominal speed. I guess Comcast is having difficulties providing enough upload capacity during daytime hours.
    I log into my DSL modem every so often and see what my down/up rates are. In QoS I put 90% of those values. That seems to be working for me obviously.
    Jim.Alles likes this.
    Untangle 16.0.1 (Build: 20201011T124207) (Kernel: 4.19.0-8-untangle-amd64)
    QOTOM-Q355G4
    1.6-2.7 GHz Intel I5 5250U, 128GB SSD mSATA, 8GB RAM DDR3L, 4xRJ-45 Intel I211AT 10/100/1000 Controller

  9. #99
    Untangler
    Join Date
    Jan 2020
    Location
    San Jose, CA
    Posts
    89

    Default

    Quote Originally Posted by theoak View Post
    I log into my DSL modem every so often and see what my down/up rates are. In QoS I put 90% of those values. That seems to be working for me obviously.
    That generally works for me as well, but it doesn't work with Cisco AnyConnect VPN. No matter how low I set the QoS value, any parallel upload will introduce severe lag on the VPN connection, even with Bypass rules =High Priority for the VPN connection active and low priority for the upload task.

    Of course the fact that upload speeds (nominal 5 Mbit/s) range from 6 Mbit/s in the evening and at night down to less than 2 Mbit/s during daytime hours doesn't help, but even with QoS upload set to 1500 kbit/s the problem occurs.

  10. #100
    Untangle Ninja
    Join Date
    May 2008
    Posts
    1,290

    Default

    This might have helped.

    https://untanglengfirewall.featureup...ncy-management

    Unfortunately it didn't into 15.1. They can't even make up their minds if it is needed. LOL

    under consideration.png

Page 10 of 12 FirstFirst ... 89101112 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

SEO by vBSEO 3.6.0 PL2