[Bloat] An interesting, and probably erroneous, article on 5G and TCP buffering

[Michael Richardson]

Dave Collier-Brown <dave.collier-brown at indexexchange.com> wrote:
    > Only being able to use 20% of the bandwidth is clearly not good (on 4G
    > the same algorithms achieve
    > 50-70%). BBR<https://blog.acolyer.org/2017/03/31/bbr-congestion-based-congestion-control/>
    > does much better with 5G, achieving 82.5% utilization. An investigation
    > reveals the problem to be caused by buffer sizes. In the radio portion
    > of the network, 5G buffer sizes are 5x 4G, but within the wired portion
    > of the network only about 2.5x (this is with a 1000 Mbps provisioned
    > cloud server). At the same time the download capacity of 5G is about 5x
    > greater: "i.e., the capacity growth is incommensurate with the buffer
    > size expansion in the wireline network." Doubling the wireline buffer
    > size would alleviate the problem. BBR does better because it is less
    > sensitive to packet loss/delay.

Is this with one flow or many thousands one would expect a real network to have?

Since we want buffers to be empty, it's unclear to me if TxOps are really
being lost, or what.


--
]               Never tell me the odds!                 | ipv6 mesh networks [
]   Michael Richardson, Sandelman Software Works        |    IoT architect   [
]     mcr at sandelman.ca  http://www.sandelman.ca/        |   ruby on rails    [

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 487 bytes
Desc: not available
URL: <https://lists.bufferbloat.net/pipermail/bloat/attachments/20201005/b71db693/attachment-0001.sig>