Procurve 2530 48G PoE stops responding over LACP trunk

[rtrbnc]

We have two switches:

1. Procurve 2530 48G PoE
2. Procurve 2810 24G

They are connected with a 2-port LACP trunk.

When I connect to http/telnet/ssh/snmp directly connected to the 2530, all is fine. However when I am connected to the 2810, I can ping the 2530 and communicate with devices connected to the 2530, but http/telnet/ssh/snmp will all initially respond once or twice, but then will all stop responding for a while. After a bit (I'd guess 5 minutes or so) it will start responding again for http/telnet/ssh/snmp, but same problem comes back and it stops responding after a few seconds. When the 2530 is not responding, I am able to access it via console, and all clients remain connected to the network and communicate across the LACP trunk without issue.

Any ideas what might cause this?

Here are the configs:

; J9772A Configuration Editor; Created on release #YA.15.12.0007
; Ver #04:01.ff.37.27:ea
hostname "HP-2530-48G-PoEP"
trunk 47-48 trk1 lacp
time daylight-time-rule continental-us-and-canada
ip default-gateway 192.168.123.254
snmp-server community "public" unrestricted
snmp-server response-source 192.168.123.241
vlan 1
   name "123 Network"
   untagged 1-46,49-52,Trk1
   ip address 192.168.123.241 255.255.255.0
   exit
spanning-tree
spanning-tree Trk1 priority 4
no tftp server
no dhcp config-file-update
password manager

; J9021A Configuration Editor; Created on release #N.11.69

hostname "ProCurve Switch 2810-24G"
snmp-server contact "xxx"
snmp-server location "xxx"
time timezone 6
trunk 23-24 Trk1 LACP
ip default-gateway 192.168.123.254
snmp-server community "public" Unrestricted
vlan 1
   name "123 Network"
   untagged 7-22,Trk1
   ip address 192.168.123.242 255.255.255.0
   no untagged 1-6
   exit
vlan 2
   name "125 Network"
   untagged 1-6
   no ip address
   exit
spanning-tree
spanning-tree Trk1 priority 4
password manager
password operator

[rtrbnc]

Update: I was mistaken that the issue does not occur when directly connected to the 2530 switch. I moved a server that was having this issue from the 2810 switch to the 2530 switch and the issue remains, despite it being directly connected to the 2530 switch.

When I do a packet capture from the server, initially the SNMP traffic flows normally, then it starts showing this over and over:

24093	13.615141000	192.168.123.235	192.168.123.241	SNMP	81	get-next-request 1.3.6.1.2.1.1
24094	13.615425000	192.168.123.241	192.168.123.235	ICMP	109	Destination unreachable (Port unreachable)