We have a network with an HP 1910-24 and a V1410-16 (unmanaged). The two are connected via two network cables, but there's no Link Aggregation configured on the 1910, and all the connections are working smoothly.

How is it supposed to work? Shouldn't it create a switch loop and make the network unusable?

Oddly enough, if we replace the 1910 with a newer device, and connect the 1410 with one (or two cables, with LACP configured), the LAN becomes unstable/unusable.

2 Answers 2


Most likely, the rapid or the multiple spanning tree protocol RSTP/MSTP has been activated on the 1910 switch.

With the unmanaged 1410 switch forwarding STP BPDUs (most do), the 1910 notices a loop to self and blocks one of the ports to the 1410. Not a really reliable state (as you're currently seeing) but mostly good to go.

if we replace the 1910 with a newer device, and connect the 1410 with one (or two cables, with LACP configured), the LAN becomes unstable/unusable.

That is to be expected. If you replace the 1910 with a switch without STP, a bridge loop forms, causing a broadcast storm to bring down the network. By default, all HPE switches I've seen come with STP turned off.

Trying to configure a LACP LAG on the new switch won't work. Lacking configuration on its 1910 partner, LACP fails and both ports are just switch ports. Again, without proper STP, a loop forms.

Creating a static LAG would work better but due to potential MAC address instabilities that's not a proper operational mode either.

In a nutshell: using anything but simple, single links between switches or more than one VLAN requires both sides to be configured in the same way: STP/RSTP/MSTP or LACP or static LAG.

If you replace any network device make sure you've examined and understood its configuration. That's where a decent documentation including the operational concept comes into play as well.

Since a device can also simply fail at any given time, make sure you've backed up its current config somewhere.

  • Great, thank you, I found MSTP enabled on the 1910. I still don't get why the new device (a Zyxel) with just one cable to the 1410 wouldn't work. Disclosure: I've never physically been there, colleagues did
    – Maxxer
    Commented Apr 27, 2023 at 11:56
  • 1
    @Maxxer If you don't know the Zyxel's config, anything can happen. For instance, BPDU guard could cause the interconnect port to be blocked. Otherwise, feel free to add the configurations to your question for us to check out.
    – Zac67
    Commented Apr 27, 2023 at 12:12
  • The Zyxel config is pretty standard, just some VLANs. Unfortunately it's now offline, so I cannot share the config. The 1410 shouldn't have a config, being it unmanaged. We deployed this switch in other networks without issues. I will retry enabling STP on the Zyxel
    – Maxxer
    Commented Apr 27, 2023 at 12:18

Most switches comes with Spanning Tree (STP) enabled by default.

This protocols will detect the loop and disable one of the link to break it.

The new switch you connect either has STP disabled or a different form of STP enabled by default, since there's different variants.

  • Thank you, the HP had indeed MSTP globally enabled. I will accept the other answer as being more detailed and comprehensive, despite having you both replied almost at the same time.
    – Maxxer
    Commented Apr 27, 2023 at 12:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.