Wired Intelligent Edge

 View Only
last person joined: 2 days ago 

Bring performance and reliability to your network with the HPE Aruba Networking Core, Aggregation, and Access layer switches. Discuss the latest features and functionality of your switching devices, and find ways to improve security across your network to bring together a mobile-first solution
Back to discussions
Expand all | Collapse all

Aruba 2930 F High Latency

This thread has been viewed 14 times

  • 1.  Aruba 2930 F High Latency

    Posted 11 days ago

    Dear Community,

    As described in the headline, this concerns an Aruba 2930F-24G-4SFP+ Switch (JL253A). I'm experiencing an issue where I encounter high latency when trying to ping, for example, 1.1.1.1 from the switch. These latency issues don't occur when I directly connect to the router and ping. I've already installed the latest firmware, version 16.11.0018. However, I've noticed that the ROM version is still running on 16.01.0010. Could there possibly be a connection here?

    Additionally, I should mention that in the event logs, I see a "Port 5 High-Collision or drop Rate" alert.

    The high latency is having a significant impact on all devices behind it, such as VOIP telephony and accessing the server via RDP, as everything runs very slowly and erratically. When the latency stabilizes again, things improve. However, it feels like the latency increases every minute.



  • 2.  RE: Aruba 2930 F High Latency

    Posted 11 days ago

    Pinging from the switch is a method to test L3 connectivity. It is not a good measure for switching or routing latency.

    Ping uses ICMP echo requests generated by the control plane, the CPU. Actual switching and routing is done by the data plane in hardware.

    And of course, pinging destinations beyond your network introduces a thousand other potential problem sources.

    "Port 5 High-Collision or drop Rate" is an indication of a problem on that port, usually a defective cable. Depending on the nature of the problem it could also cause high latency on the path. Check error counters on both sides of the link, optical power levels etc to diagnose.

    And yes, you should keep the firmware somewhat up to date.


    Original Message


  • 3.  RE: Aruba 2930 F High Latency

    Posted 6 days ago

    Hello,

    Thank you for the prompt response! I will definitely replace the cable on Port 5. In the meantime, I have looked into the details of the SFP modules and obtained the following results.

    vfg1-sw1# show interface transceiver detail

    Transceiver in 25
       Interface Index    : 25
       Type               : SFP+SR
       Model              : J9150A
       Connector Type     : LC
       Wavelength         : 850nm
       Transfer Distance  : 80m (50um), 30m (62.5um), 300m (50um OM3),
       Diagnostic Support : DOM
       Serial Number      : G1908062030

     Status
       Temperature : 33.328C
       Voltage     : 3.3311V
       Tx Bias     : 6.090mA
       Tx Power    : 0.4951mW, -3.053dBm
       Rx Power    : 0.5065mW, -2.954dBm

       Wavelength         : 850nm
       Interface Index    : 26
       Type               : SFP+SR
       Model              : J9150A
       Connector Type     : LC
       Transfer Distance  : 80m (50um), 30m (62.5um), 300m (50um OM3),
       Diagnostic Support : DOM
       Serial Number      : G1908062031

     Status
       Temperature : 34.054C
       Voltage     : 3.3358V
       Tx Bias     : 6.082mA
       Tx Power    : 0.5666mW, -2.467dBm
       Rx Power    : 0.6769mW, -1.695dBm

    Transceiver in 27

       Interface Index    : 27
       Type               : SFP+SR
       Model              : J9150A
       Connector Type     : LC
       Wavelength         : 850nm
       Transfer Distance  : 80m (50um), 30m (62.5um), 300m (50um OM3),
       Diagnostic Support : DOM
       Serial Number      : G1908062029

     Status
       Temperature : 34.984C
       Voltage     : 3.3311V
       Tx Bias     : 6.142mA
       Tx Power    : 0.5587mW, -2.528dBm
       Rx Power    : 0.7588mW, -1.199dBm

    Transceiver in 28*
       Interface Index    : 28
       Type               : 1000SX
       Model              :   ??
       Connector Type     : LC
       Wavelength         : 850nm
       Transfer Distance  : 550m (50um), 300m (62.5um),
       Diagnostic Support : DOM
       Serial Number      : unsupported

     Status
       Temperature : 33.000C
       Voltage     : 3.2910V
       Tx Bias     : 6.204mA
       Tx Power    : 0.5826mW, -2.346dBm
       Rx Power    : 0.5153mW, -2.879dBm
    * third-party transceiver


    I think i might switch out the transceiver for a supported one 


    Original Message


  • 4.  RE: Aruba 2930 F High Latency

    Posted 6 days ago

    This is obtained from the switch System Monitor 


    Original Message


  • 5.  RE: Aruba 2930 F High Latency

    MVP GURU
    Posted 10 days ago
    ArubaOS-Switch Software version and Boot ROM version are two different things (even if their naming - versioning - looks like quite similar): check their versions using show version and show flash commands, Boot ROM is generally tied to (and, when necessary, automatically updated with) the Software version used.

    Using ICMP Ping to measure the "responsiveness" of a network while two peers (are both local? is one local and the other remote? is routing involved? is a next hop involved?) communicate through it could be misleading even if the network is running smoothly (no issues, no over-provisioned/over-loaded, stable, etc.).

    Testing from (or to) a Switch control plane could be misleading too (for the reasons already well suggested in the first reply to your post).

    A test you can do to understand if there are internal issues is to measure average ICMP Ping time and Packets Loss (if any) between two locally connected peers (consider that if those peers belong to different VLANs and those VLANs are routed by your Firewall instead of being routed by your routing switch there will be a delay introduced by your Firewall...).

    Other test should use IPERF (to load involved interfaces). Clearly used peers should be well configured to avoid bottlenecks or traffic filtering.

    The above could be useful to stress your network looking for specific issues...given that basic ones (like defective cabling, packet storms, interface misconfiguration, routing issues, etc.) can already be excluded since beginning.

    You're testing ICMP Ping from Switch console to a public DNS (!!) versus testing ICMP Ping from your next hop gateway, the Firewall (doing Routing and NAT), to a public DNS (!)...well, if so, even forgetting that most of the latency is accounted by the packets' journey on the wild (outside your Firewall), you should then check how your switch is connected to its gateway (physical uplink and logical uplink).



    Original Message