HP 1920 temporary no reply to ICMP request

[Marco2]

I am observing a temporary not reply to icmp request on different switches 1920. Traffic pass through correctly but, for little time, switch not reply to ping request. In fact, IP is associated to Vlan Interface and I noted the issue using ping -t (the not reply could be appears only one time at day). When switch not reply, in the arp table is not present, and after some seconds it appears again and start to reply to ICMP. I tried from some PC and updating firmware not fixed issue. In my network are present many HP switch (2610, 2530, 5500, 1920..) and problem is only on 1920. In lab I could not simulate the issue, switch works properly, connecting this switch in my broad network issue returns.

Any suggestions?

Many thanks in advance

Marco

[TerjeAFK]

We don't have any 1920 switches here, but I have seen this on low-end HP switches before (1800 series for example) where the switch has been under high load and has dropped ICMP packets for a short time. Can you check CPU and memory on the switch when this happens?

[Marco2]

Thank you for suggestion. CPU is stable around 4% and memory is for 42% used. Issue seems not correlated to memory or cpu high load.

[parnassus]

Could it be something related to (how fast) ARP Table fills up or related to the ARP aging timer (on the HPE 1920 Switch Series the default arp timer aging is 20 minutes and it can range between 1 and 1440 minutes)?

[Marco2]

Thank you for suggestion. I already tried to change arp entry from dynamic to static for server monitor but issue was still present.

[Burg26]

Same problem here.

Appeared to start having this problem after enabling Aruba Central.
Latest Firmware.

Somebody found a solution?