Troubleshooting with Marvis – Missing VLAN

Troubleshooting with Marvis – Missing VLAN

This post is part of the series, “Troublehsooting with Marvis”. I decided to write these to document some connectivity related troubleshooting tips using MistAI and Marvis. When it comes to any WLAN, troubleshooting the network post deployment is an intrinsic part of any WLAN engineers job.

User Satisfaction and Performance:

Users are crucial. Their experience during the Wi-Fi connectivity phase plays a vital role in any WLAN deployment. When users begin to encounter problems while connecting to a wireless network, solving their problems quickly becomes of paramount importance.

All WLAN vendors provide different tools for connectivity troubleshooting. Simply put, MistAI and Marvis make troubleshooting connectivity issues easy. “My Wi-Fi is not working is a common term that all WLAN engineers are used to hearing; regardless of what the issue is.

What to do?

We can not simply say, “It is not the Wi-Fi” and go back to doing something else. For an MSP, first call resolution is crucial to customer satisfaction and business needs. Each escalation and time to resolve an issue costs the company significant money. Our goals should be:

  • Proactively fix the issue.
  • When customer calls; find the root cause of the issue.
  • Resolve the issue.

This is where a well designed interface to Deploy and Troubleshoot a Network and a WLAN environment comes into play.

Mist AI and Marvis:

Mist AI and Marvis play a vital role in proactively monitoring and troubleshooting issues. Many common issues can be solved before customer picks up the phone and calls. This reduces help desk calls, increases customer satisfaction and minimizes escalations.

Topology:

Let’s take a simple WLAN topology, 802.1X SSID with multiple user groups. Each user group has a different VLAN attribute. Any user belonging to a certain user group gets assigned that specific VLAN.

WLAN – eapsec

SSID “eapsec” is being served by three different VLANs, 20, 30 and 40. I have purposely removed VLAN40 from the trunk port that Mist AP43 is connected to. This scenario results in a very common problem when client connection is successful but they are unable to grab an IP address (Note: there are other reasons for not getting an IP address also; we are only discuss missing VLAN).

Client Properties
iPhone

Client properties and iPhone image show that the client connected to the SSID successfully. Machine state and EAP Authentication was successful; yet client does not have a valid IP address.

Frame Capture

Here you can see the client sending out the broadcast for DHCP with no success

Reactive (Legacy WLAN):

With legacy WLAN deployments, a simple issue like this can take some time to resolve. I’ve added a very basic flow of the process.

Call Flow

Looking at the call flow, it is clear that a simple missing VLAN from a single trunk port of an AP can cause an outage for multiple WLAN users. With the reactive approach, it can take some time to resolve the issue.

Proactive (MistAI, Marvis):

With a Juniper-Mist WLAN deployment this issue can be resolved proactively. Let’s see how MistAI and Marvis will handle this.

E Mail Notification:

After I setup this topology within a few seconds this email notification was received in my email. E Mail notifications can be setup for multiple emails. These can go to individual email boxes or a general help desk monitored email. As soon as help desk agent(s) see this email, they can create a ticket and resolve the issue before customer even realizes that there is an issue.

Missing VLAN – E Mail notification
Missing VLAN – E Mail notification

Wireless SLEs:

Once logged into the dashboard, wireless SLE’s will proactively show the ongoing issues. Follow two SLE’s “Time to Connect” and Successful Connects” show “DHCP” issues. These can be further investigated by clicking on the “Sub-Classifiers”.

Time to Connect
Successful Connects

From here I can dig deeper into the issue by clicking on “Affected Items”. This will show me all the devices and general area that is being affected by this outage.

Affected Items

I can click on the client and access “Client Insights”. Here I can see all the client events and also look at the frame captures for further investigation and resolving the issue.

Client Insights
Client Events

Red Alerts:

Once you click on the switches you will be presented with a high level overview of what is going on with the switches. One of the metric is “VLANs”.

So now I know that there is an issue with a VLAN; but where? After I click on the VLANs box, I can see exactly which access point and which port is experiencing this issue. Helpdesk, NoC, TAC or whoever handles the monitoring of the environment can easily fix this issue in seconds from here and document it in the ticket. Success can be validated by calling the users in that area or simply looking at the clients that were previously affected.

Marvis Actions:

This is another method that allows first level support to proactively fix an issue before they start receiving customer calls. Marvis Actions dashboard here is showing me that there is a missing VLAN issue with a switch.

Marvis Actions

I can click on that Marvis Action Alert and find out which switch is having this issue.

This is still not enough to fix/resolve the issue, I can now click on “View More” and voila. Now I know the switch and the AP that are part of the problem. I know which VLAN is missing and from which port. It can easily be added to that switch port now and issue can be resolved in a matter of minutes.

Summary:

Wi-Fi access has become a crucial part of our lives. Users expect it to work. With a drastic increase in users utilizing Wi-Fi as their primary method of network access, and mission-critical application access over Wi-Fi. Gone are the days of fixing a Wi-Fi related issue in hours. SLAs that used to be hours are now minutes. Mist AI and Marvis help alleviate the issues faced by legacy WLAN solutions.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

WordPress SEO