frequent Disconnections

[IT-woman]

We have Meraki CW9166I AP's at one of our sites that keeps losing connection or rebooting. Some of the disconnects say "Disabled gateway(bad DNS)"

We have other sites that are all configured the same way, with no issues. Can you tell me what could be causing the error?

[DarrenOC]

Hi @IT-woman , with regards the DNS error have you verified that the required ports are allowed on your upstream firewall?

Are devices that are using the APs losing connection or is it just the APs losing connectivity to the cloud?

[IT-woman]

DNS server for that particular site was pointing to the DC at our main location. We have switched it to use the local DC for DNS. This may have resolved the issue.

[DarrenOC]

Fingers crossed @IT-woman .  As they always say “it’s always DNS”

[DarrenOC]

 

[IT-woman]

Hello @DarrenOC  So the issue turned out to be a CSCswbug 18530. The workaround we used is below, since adding the fix we haven't had any other disconnection.

Aug 30 14:06:28.620: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet1/0/22, changed state to upAug 30 15:14:54.218: %ILPOWER-3-CONTROLLER_PORT_ERR: Controller port error, Interface Gi1/0/22: Power Controller reports power Imax error detectedAug 30 15:14:55.658: %LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet1/0/22, changed state to downAug 30 15:14:56.661: %LINK-3-UPDOWN: Interface GigabitEthernet1/0/22, changed state to down 

Regarding to the show tech of switch2, the information was incomplete. However, we think that this device is showing similar logs related to the Gig1/0/14 interface.

According to these logs, these devices could be affected by CSCsw18530 bug. In the following link you can find more details:

https://bst.cisco.com/bugsearch/bug/CSCsw18530

In this case, the following are the suggested workarounds to try to solve this issue. If the first workaround doesn’t work, please continue with the next one and so on:

• To use a longer cable (>50ft) between devices seems to fix the issue in most cases.
• Apply the following commands in the Gig1/0/22 – SW3 and Gig1/0/14 – SW2 interfaces:

#conf term

(config)#int x/y/z

(config-if)#power inline port 2x-mode

(config-if)#shut

(config-if)#no shut

(config)#exit

• Configure static power these interfaces (Gig1/0/22 and Gig1/0/14):

#conf t

(config)#interface x/y/z

(config-if)#power inline static max 30000

(config-if)#shut

(config-if)#no shut

(config)#exit

[DarrenOC]

Excellent news @IT-woman .  Thank you for sharing the resolution with the community.

[IT-woman]

Hello DarrenOC

Yes, all ports are allowed on our FW.

Devices using the APs are losing connection and I think the APs are losing connection to the cloud. How do I determine if the APs are losing connection from the cloud?

A few of the errors show the AP rebooted, and some say "Disabled gateway bad DNS", others say "No connectivity" Then they'll reconnect,  It has been intermittenly happening since the beginning of this month.