GRitchie Posted March 15, 2016 Posted March 15, 2016 I've got a support case open with HP for this, but wanted to ask here too: I've got a problem with some HP kit (brand new HP 2530 switch). When using the ports as an uplink to another switch (NETGEAR - no fibre SFP slot) the port on the HP switch disables itself after so long. I may not have worded that brilliantly, so ask away if need be. Really need to get to the bottom of this!
Geoff Posted March 15, 2016 Posted March 15, 2016 Spanning Tree, BDPU guard or loop protection can cause this. What does the logging say?
GRitchie Posted March 15, 2016 Author Posted March 15, 2016 Have got Spanning Tree and (I think) loop-protect enabled on these. The logging says nothing about that port :/
Geoff Posted March 15, 2016 Posted March 15, 2016 It should say if it is disabling it in the logging.
GRitchie Posted March 15, 2016 Author Posted March 15, 2016 Had a look in show log -r but nothing is there...
Michael Posted March 15, 2016 Posted March 15, 2016 So are you using an SFP on the HP end and a Media Converter on the Netgear, or literally Ethernet to Ethernet? If Ethernet to Ethernet, I presume you haven't forced the mode? I can't see why Spanning Tree would stop the port working, most likely a speed/link issue, or possibly a cable issue?
GRitchie Posted March 15, 2016 Author Posted March 15, 2016 I've got an SFP on the HP switch, into the fibre panel. I've got a Netgear switch going from (for example) Port 5 on the HP switch to an Ethernet uplink port (on the netgear). I've not forced anything, not sure what you mean there. Nothing wrong with the cable, swapped that out and I've currently just swapped two of the switches going into the HP switch around, so more of the network is on than off.
mowgli82 Posted March 15, 2016 Posted March 15, 2016 I've seen Cisco switches disable ports due to excessive packet loss or jabber - we had a remote building connected by laser link and when it snowed or rained heavily the switches would disable the problem ports.
GRitchie Posted March 15, 2016 Author Posted March 15, 2016 I see. There was some mention of: Excessive CRC/% alignment errors and something to do with jabber (which seems to have disappeared from the logs???) but not on the ports I'm having problems with.
pantscat Posted March 15, 2016 Posted March 15, 2016 Is there newer firmware available for the switches? Generally speaking that's the first thing that HP would get you to do...
Michael Posted March 15, 2016 Posted March 15, 2016 So if I've understood correctly, the SFP is linking off to somewhere else and that this is an Ethernet to Ethernet issue. I'd tried literally connecting the two switches together with no specific configurations and see if the issue goes away.
Michael Posted March 15, 2016 Posted March 15, 2016 By force, I meant specify 10, 100 or 1000 instead of Auto, which is normally the default.
GRitchie Posted March 15, 2016 Author Posted March 15, 2016 (edited) Don't think there's a newer firmware. Updated these quite recently. @Michael - SFP to fibre patch panel (to server room). Netgear to HP switch is eth-eth (which is the problematic connection) yes, that's correct. As these switches are in production now, or at least one (the working part of the HP switch) I can't necessarily take them out to test/format etc. unfortunately - without a decent amount of downtime. UPDATE: Yes, they're all on auto Edited March 15, 2016 by GRitchie
Michael Posted March 15, 2016 Posted March 15, 2016 If you've tried swapping the cable, I'd suspect it's a possible config issue. As I say, just try connecting the two together with no configuration on either port (both HP and Netgear). It should work, as at this level it's just using Ethernet and no specific protocols.
GRitchie Posted March 15, 2016 Author Posted March 15, 2016 That's what it is now though... I've just done an upgrade which apparently I hadn't done (mind the logo has now changed from HP to Aruba??)
Michael Posted March 15, 2016 Posted March 15, 2016 I guess you could try Auto-100 each end and see if this makes a difference. I appreciate not ideal as it creates a bottleneck, but worth seeing if the connection is kept live longer.
twin--turbo Posted March 15, 2016 Posted March 15, 2016 Put both ends to 100-half see if its stable. Then 100-full , then 1000{duplex is irrelevant} Also download the free kiwi syslog and set both switches to log debug to kiwi, much easier to spot errors that way. TT Ps. Are these the same ones that were causing errors a few months back? 1
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now