DHCP Server Unreachable Errors

  • 1
  • Question
  • Updated 2 years ago
  • (Edited)
We seem to be running into some DHCP issues and dropping clients even after the "sip alg enable" fix for forwarding tables filling up.  The problem seems more localized to one school now in one part of the building.  Here's the problems I'm seeing:

1. About 450 "DHCP Server Unreachable" errors in a 3 hour period.  In that period, we had 1157 unique clients connect.  Almost all errors are on our Public which I outlined how this is setup further down in this post.

2. The APs seem to be stopping all wireless communication at times.  Here's what I'm seeing in PRTG.  This graph tracks TX Airtime on the 5GHz radio:


Here's the corresponding Connected Clients graph in HiveManager NG during the same time period:


3. Our tech on-site has reported multiple occurrences of clients losing connectivity only to be left with a 0.0.0.0 IP address on both the LAN and Public SSIDs.
 

Here's our router and switch setup:

1. FortiGate with two VLANs setup; VLAN 1 acting as the LAN, VLAN 10 acting as the Public.

2. Windows DHCP server living on VLAN 1.  FortiGate does DHCP relaying from VLAN 10 to the DHCP server on VLAN 1.

3. HP switches with VLAN 1 untagged on all ports and VLAN 10 tagged on all ports.  Access points have a separate SSID for each VLAN.

4. Clients mostly consist of Chromebooks with 802.11AC wireless cards and iOS/Android devices connecting to the Public.


I also attached the tech data and a 2 minute packet capture on the 5GHz radio interface on a problem access point.  Let me know if you need anymore information from me!

Packet Capture: https://drive.google.com/file/d/0B4tISa9xO-7kYjQ4Y3IzSWF4SHM/view?usp=sharing

Tech Data: https://drive.google.com/file/d/0B4tISa9xO-7kM05Cem5QQXhsZzA/view?usp=sharing
Photo of Devin

Devin

  • 17 Posts
  • 1 Reply Like

Posted 2 years ago

  • 1
Photo of Kevin Gee

Kevin Gee

  • 54 Posts
  • 4 Reply Likes
Hi,

I've encountered similar problems in the past with AP 230s but that may have been fixed in recent HiveOS versions.If it's the problem I'm thinking of the immediate ugly fix was rebooting the AP. I think at the time one of the recommendations was to disable QoS is you have any set on the APs.

I would be helpful to specify what model APs are involved and what HiveOS version is running.

Regards, Kevin.
Photo of Devin

Devin

  • 17 Posts
  • 1 Reply Like
I apologize for not including that in my first post.  We have 56 AP250s running 7.0r2 at that school.  I do have "Dynamic Airtime Scheduling" enabled if that's a known bug.  Otherwise, we've done multiple reboots, but the problem pops up multiple times throughout the day on multiple APs so that isn't too reliable of a solution for us.
Photo of Luke Harris

Luke Harris

  • 265 Posts
  • 18 Reply Likes
I would also look at QoS Airtime settings as well. I have disabled this within my deployment to reduce CPU overheads. 
Photo of Matthew Norwood

Matthew Norwood, Employee

  • 4 Posts
  • 0 Reply Likes
Devin,

This issue was fixed in HiveOS version 7.1r1 which is generally available now. Hopefully that fixes your problem.

Link to the release notes here: http://docs.aerohive.com/330000/docs/help/english/documentation/7.1r1_HiveOS_ReleaseNotes.pdf

Matthew