Instant On - Wired

Hi everyone!

I recently bought the Aruba Instant On 1960 12XGT 4SFP+ mainly for its 10G ports. I use 2 SFP ports in LACP as downlinks to another switch, ports are running in 1G until I can replace the other switch.

I use the other 2 SFP ports in 10G mode, LACP for a NAS.

2 RJ45 ports are LACP to a 3rd switch.

Ports 1 to 5 on the Aruba are used for a Kubernetes cluster, all 5 are 1gbps. The issue is with these 5 ports, they seem to go down and back up continuously.

Initially, I had enabled flow control, storm control and jumbo frames. There were no issues until I connected the 5 kubernetes nodes. I disabled flow, storm and jumbo frames first thing, but they are still going down and up continually. I have attached a part of the log and interface statistics.

I checked the hosts kernel logs etc, but they are not showing any issue.

When I connect the hosts to the my original switches (Ubiquiti Unifi) I see no errors or problems whatsoever. Flow control is enabled there as well, jumbo frames not and I don't think it has storm control settings.

I use Netdata to monitor Kubernetes and I'm getting spammed with unreachable notifications. Literally hundreds for the past day.

I had also logged a support case which actually worked very well! I immediately received a response which asked for more information and generate a support file. The responses were fast and  a few hours later they suggested to have a Teams call. Shared the screen with the Aruba Web interface and we went over the settings. She suggested a few changes and monitor over the weekend:

Spanning Tree / Global / Spanning Tree Admin Mode: Disabled
Switching / EEE Configuration / Global / Auto Port Power Down: Disabled
Switching / EEE Configuration / Global / Low Power Idle (EEE): Disabled
Switching / Loop Protection / Global / Loop Protection: Enabled
Switching / Loop Protection / Interface Configuration: All Interfaces: Loop Protection: Enabled

Since I had been receiving dozens of alarms for my nodes becoming unreachable and reachable again per hour, it was quite fast to conclude that one of these settings was the culprit. I have received no alarms since yesterday 16:00.

I don't expect STP to be the one, I think its the energy settings that was disabling ports. I could be wrong of course, its just a hunch.

She suggested to create another support file monday, share it and we will narrow down what the issue was.

Original Message:
Sent: 04-25-2024 06:55 AM
From: Shuttervalve
Subject: ports going down and up

Hi everyone!

I recently bought the Aruba Instant On 1960 12XGT 4SFP+ mainly for its 10G ports. I use 2 SFP ports in LACP as downlinks to another switch, ports are running in 1G until I can replace the other switch.

I use the other 2 SFP ports in 10G mode, LACP for a NAS.

2 RJ45 ports are LACP to a 3rd switch.

Ports 1 to 5 on the Aruba are used for a Kubernetes cluster, all 5 are 1gbps. The issue is with these 5 ports, they seem to go down and back up continuously.

Initially, I had enabled flow control, storm control and jumbo frames. There were no issues until I connected the 5 kubernetes nodes. I disabled flow, storm and jumbo frames first thing, but they are still going down and up continually. I have attached a part of the log and interface statistics.

I checked the hosts kernel logs etc, but they are not showing any issue.

When I connect the hosts to the my original switches (Ubiquiti Unifi) I see no errors or problems whatsoever. Flow control is enabled there as well, jumbo frames not and I don't think it has storm control settings.

I use Netdata to monitor Kubernetes and I'm getting spammed with unreachable notifications. Literally hundreds for the past day.

I don't understand what these packets with errors are or why that's happening.  Port 10 is my laptop.

I replaced the UTP cables (they are brand new, good quality). While it could be a cable issue, I doubt its all 5 and they work fine on other devices and with the Unifi switches.

Can help me out here to figure out what is wrong? I was going to replace all my Unifi switches with these so I can have 10G links between them but I do really need to fix this first.

Thanks in advance!

ps. lve been reading the user manual and been searching for 2 less important questions:

- ssh access; how do I enable that? I wanted to see if it would provide better logging / troubleshooting

- webgui timeout increase, is this possible? Its infuriatingly short, in particular when troubleshooting, going back and forth between windows.

I am having this exact same problem.  It appears that no matter how you set your port and stp settings, port status will appear in log as warning regarless.  I have been working with HPE over the past week or so, just had a remote session with escalated engineer as I was having no luck with any suggestions they intially provided.  After all emails back and forth a remote session, it was determined that there is no fix for this.  According to their engineer, this was problem was done intentionally.  Obviously, he was trying to mask the fact that they made a massive mistake and refuse to correct it.  Edge ports should not be showing constant up and down notifications every time a pc is used.  These switches are garbage. 

Original Message:
Sent: 04-25-2024 06:55 AM
From: Shuttervalve
Subject: ports going down and up

Hi everyone!

I recently bought the Aruba Instant On 1960 12XGT 4SFP+ mainly for its 10G ports. I use 2 SFP ports in LACP as downlinks to another switch, ports are running in 1G until I can replace the other switch.

I use the other 2 SFP ports in 10G mode, LACP for a NAS.

2 RJ45 ports are LACP to a 3rd switch.

Ports 1 to 5 on the Aruba are used for a Kubernetes cluster, all 5 are 1gbps. The issue is with these 5 ports, they seem to go down and back up continuously.

Initially, I had enabled flow control, storm control and jumbo frames. There were no issues until I connected the 5 kubernetes nodes. I disabled flow, storm and jumbo frames first thing, but they are still going down and up continually. I have attached a part of the log and interface statistics.

I checked the hosts kernel logs etc, but they are not showing any issue.

When I connect the hosts to the my original switches (Ubiquiti Unifi) I see no errors or problems whatsoever. Flow control is enabled there as well, jumbo frames not and I don't think it has storm control settings.

I use Netdata to monitor Kubernetes and I'm getting spammed with unreachable notifications. Literally hundreds for the past day.

I don't understand what these packets with errors are or why that's happening.  Port 10 is my laptop.

I replaced the UTP cables (they are brand new, good quality). While it could be a cable issue, I doubt its all 5 and they work fine on other devices and with the Unifi switches.

Can help me out here to figure out what is wrong? I was going to replace all my Unifi switches with these so I can have 10G links between them but I do really need to fix this first.

Thanks in advance!

ps. lve been reading the user manual and been searching for 2 less important questions:

- ssh access; how do I enable that? I wanted to see if it would provide better logging / troubleshooting

- webgui timeout increase, is this possible? Its infuriatingly short, in particular when troubleshooting, going back and forth between windows.