Network Management

 View Only
last person joined: 6 days ago 

Keep an informative eye on your network with HPE Aruba Networking network management solutions

Broadcast storm and port blocked with Ubiquiti U6-LR

This thread has been viewed 5 times
  • 1.  Broadcast storm and port blocked with Ubiquiti U6-LR

    Posted Aug 01, 2022 10:03 AM
    Hi guys, I have a somewhat anomalous problem that I honestly don't know how to fix: I have an infrastructure consisting of about twenty Aruba HP 2530 switches (8, 24 and 48) to which they are connected, distributed throughout the building, 50 Ubiquiti U6-LR.
    The access points are driven by UniFi Network Application 7.1.68, installed on a linux VM.
    Generally everything works without problems, the APs update quietly, even Network Application updates from the command line without problems.
    But, sometimes, I run into a broadcast storm that is intercepted by the spanning tree protocol of the switches that disables the ports ( to which the APs are individually connected) or blocks the access points (the APs no longer receive the IP via DHCP, managed upstream by the firewall, or they lose the pairing with the controller), also interrupting the functionality of the wifi network as if a loop were blocked, which in reality (apparently) there is not ...
    In that case I have to individually restart the switches affected by this problem, restart the vm and wait for the problem to disappear.
    I cannot understand if this is due to a problem with the STP protocol (RPVST mode) or some misconfigured Network Application parameter.
    For the sake of completeness, I use multiple VLANs in the network: the one used for the IPs of the access points is ID100 (untagged on the switches), while the ID20 is the one responsible for issuing the IPs for devices connected to wi-fi (tagged on the switches).
    When this happens, the errors that are returned are "STP - VLAN 100 - Starved for a BPDU on port x from Designated Bridge" and "port x - Excessive Multicast".
    After a while I noticed that this anomalous behavior is triggered either by the restart of the switches or, above all, by the restart of the vm and that therefore, having the APs to realign with the controller, they send an excessive broadcast that clogs the network.
    I could turn off spanning tree, but honestly I would avoid ...
    Do you have any idea how to solve the problem?

    ------------------------------
    Leviathan
    ------------------------------