Meeting 2024 03 - freifunk-gluon/gluon GitHub Wiki

Gluon Meetup 2024-03

11.06.2024 - 20:00 CET - mumble.freifunk.net

Notes from Last Meetup: Meeting-2024-02

Attendees

  • [ :bust_in_silhouette: djfe]
  • [ :bust_in_silhouette: joejelliss]
  • [ :bust_in_silhouette: neocturne]
  • [ :bust_in_silhouette: tomh]
  • [ :bust_in_silhouette: T_X]

Topics

Please always note your name next to your topic

  • [ :bust_in_silhouette: djfe] Ubiquiti Airmax

    • #3141
    • The PR hasn't been tested since February. Some of the tests led to the device needing to be recovered.
    • djfe knows an owner of a Nanostation XM but the owner currently doesn't have time for Freifunk either.
    • The PR aimed to introduce a more specific way to controll the flag that fixed it and not change something for all devices of the target.

  • [ :bust_in_silhouette: djfe] Wireguard VPN doesn't set mtu to value defined in site.conf

    • #3258
    • The wireguard mtu shouldn't matter due to the vxlan(?) interface already having a lower mtu. The 1420 should never be reached
    • Neocturne voiced interest in simplifying our MTU documentation https://gluon.readthedocs.io/en/latest/user/mtu.html#maximum-mtu
    • discussion on potential ways to automize MTU detection -> tricky
      • T_X: there is an RFC for MTU propagation on/between tunnels though: https://www.rfc-editor.org/rfc/rfc4459#section-3.2 (not implemented in batman-adv/fastd/l2tp/wireguard/vxlan)
      • This wouldn't solve that one-to-many VPN taps can only offer one MTU for all devices.
      • T_X: it would allow setting high MTUs on all tunnel interfaces though and signaling/propagating lower MTUs to other, lower encapsulating protocols and down to the final clients via ICMP(v6)

  • [ :bust_in_silhouette: djfe] ath10k issues on v2023.2.x that weren't present on v2023.1.x

    • ath10k_pci 0000:00:00.0: failed to flush transmit queue (skip 0 ar-state 1): 0
    • Noone remembers this issue, it might be new.
    • It was recommended to open issues on Github to track these (in Gluon as well as upstream in OpenWrt).
    • Maybe rolling back the ath10k firmware to an earlier state fixes this issue.

  • [ :bust_in_silhouette: djfe] ath9k issues on v2023.2.x that weren't present on v2023.1.x

    • ath: phy0: unsupported hw bitrate detected 0x1b using 1 Mbit
    • read the topic above

  • [ :bust_in_silhouette: djfe] documentation PRs

    • Feel free to review
    • There is a Github Action that generates a Read the Docs for reviewing: Click "Details" next to the last action "docs/readthedocs.org:gluon — Read the Docs build succeeded!"
    • #3188
    • #3267