How to fix high ping problem?

Question

linux

Hello! The bottom line is
there is ESXi, it has a mail server, a gateway (kerio control), several other machines and a newly raised Wi-Fi controller from UniFi (debian 7.7 x86_64).
Everything works on one vlan, except for wi-fi (it hangs on vlan10: it is thrown on switches and comes to the gateway). Today (uptime is 40 days, before that everything was ok, nothing has changed) high ping periodically falls from the main network to the controller (up to 700ms), it lasts from a minute to 10. The access points themselves ping perfectly. both the gateway and the controller feel great.
What I have already done:
1. Limit the bandwidth on the vlan10 interface on ESXi (there were suspicions that the network card could not cope with mail, gateway and controller at once) - drop packets disappeared. ping also appears high
2. Disabled Ipv6
Inside the controller itself:
a bunch of mongodb + java-6-jre + unifi
all statistics are excellent (top, iostat, ifstat, ifconfig, sar) - there are no limits anywhere.
what could be the problem or how to diagnose?

Answer 1

What I already did:
...
2. Disabled Ipv6

What does it mean? Where was it turned off? On every workstation on the network?

all stats are great (top,iostat,ifstat,ifconfig,sar) - there are no limits anywhere.

Did you look at the indicators on the switch ports? Number of dropped packets, queue sizes?

what could be the problem or how to diagnose?

Alternatively, broadcast/multicast traffic peaks. Diagnose by removing relevant data from switch ports or by analyzing traffic.

Answer 2

Ping from the point to the device is normal at the same time? Suffering ping from point to controller?