Schedule reboot or alert administrator when client encounter problem connecting to AP

  • 1
  • Question
  • Updated 4 years ago
  • Answered
I have a customer currently using Buffalo, Dlink and Ruckus APs in their network. Just setup a POC for them to test out Aerohive. One of the requirement is that they need the AP to be capable of doing schedule reboot. Reason been that they have cases whereby clients start having problem connecting to the AP after approximately 1 month later. Can I know if I can configure that on my HMOL, or if there's anyway to alert the administrator when client is having issues connecting to Aerohive APs.

Just to share with what customer has told me. Buffalo has the feature to reboot their AP (Not surprising consider they are more of a consumer grade AP). However, what's surprising is that Ruckus has the same problem with client's connectivity, and it's a higher end model (ZoneFlex 7962).

Regards
Kevin
Photo of Kevin Chen

Kevin Chen

  • 4 Posts
  • 0 Reply Likes

Posted 5 years ago

  • 1
Photo of Tash Hepting

Tash Hepting

  • 55 Posts
  • 29 Reply Likes
Kevin,

While we do have the ability schedule a reboot at a future day/time on the CLI, we do not currently have the option to make it recurring. Hivemanager also allows you to reboot groups of APs (location or tag based) as a single operation, but again not in a scheduled or recurring manner.

Generally, we try to focus on resolving the underlying issue rather than a service-disrupting bandaid (automatic reboots are not appropriate for many enterprise deployments). From past experience, client behavior is so varied that it unfortunately ends up having to be addressed on a case-by-case basis for clients that aren't from the major vendors. The standards provide a good framework for interoperability, but they still leave a little bit of room for interpretation and it's extremely difficult to predict what everyone will do.

Do you have any more detail on the client, or the nature of the connectivity issue? There are a lot of ways that the client could fail to connect, and while there is a lot of logging and alerting you can turn on I'm not sure that you'll be able to tailor it to a catch a specific client issue without also including a lot of unneeded information.

Regards,
Tash
Photo of Crowdie

Crowdie, Champ

  • 972 Posts
  • 272 Reply Likes

I am with Tash on this.  

We have a number of large educational deployments, with some students having four of five devices connected, and we do not have an issue with needing to reboot the Aerohive AP330 and AP350 access points.  These deployments are generally high density so the access points can be under some heavy load for long periods of time.  However, we have experienced some issues with the AP121 access points when they are put under heavy load for long periods of time so I recommend utilising the AP330 and AP350 for deployments that will come under heavy load (unless you want to deploy 802.11ac access points).

If you are going to deploy a high density wireless network the 802.11b data rates must be disabled (otherwise the channel utilisation will be extremely high causing performance issues) and keep the transmit powers low.  How low the transmit power needs to be should be determined during the site survey and design phase of the deployment.

There is no surprise that the D-Link access points need to be regularly rebooted.  This is the difference between an enterprise grade access point and a SME grade access point.

Photo of Bradley Chambers

Bradley Chambers, Champ

  • 302 Posts
  • 53 Reply Likes
I ran into something similar the other day. We just rolled out OpenDNS Umbrella on out laptops and one of my users was on another campus that used Cisco APs. The user had a tough time getting online and best I can tell it was because their network was forcing a certain DNS while mine Umbrella client was forcing OpenDNS.

These are always tough issues because it can be a ton of things (from the actual client to the switch). One issue I ran into is a user was complaining that our WiFi wasn't good to a few other employees, but when I inquired about it, he fessed up that is was my content filter was blocking him from surfing YouTube while at work.
Photo of Jason Hills

Jason Hills

  • 78 Posts
  • 3 Reply Likes
on a similar but different note...
Is it possible to schedule a upload of the config to the APs?  

It's not normally appropriate, but In this case we have a customer who wants to use Usermanager to add now 'Permanent' PPSK users, but has not rights to login to perform a conifg update.They are not keen on using assigning new users to the 'auto' PPSKs already available on the APs.
Scheduling a regular update would solve this problem. Providing nobody else has any pending changes awaiting upload.

thanks...