Network Management

So we've recently taken on board an Aruba wifi setup which is new to us as we have always been Cisco only before in regards to APs and management of...

I've read around the boards after a few seraches and I've seen various responses that say "you don't need to monitor" the APs via ping, etc as it is all managed via the controllers. I understand that but we have a scenario that I haven't found an answer to on the boards (maybe the wrong search terms??)...

So, our setup: We have 2 x Aruba 7240's in a Master / Backup setup with 4 further 7240's as controllers of the >2,100 AP's. The AP's are in campus mode if that makes a difference and the WLC's are all running AOS 6.5.0.3

We need to be able to proactively monitor the AP's via our current monitoring platform (MP) which integrates directly with our call management system (CMS). Without sigificant development and cost we can't integrate the WLC's with the CMS so we need to be able to monitor basic up / down of the AP's.

How can we achieve this?

If it makes a difference (or helps) then the MP is WhatsUp Gold v16 and

CMS is BMC.

Thank you.

If you can ask Whats up Gold, what kind of support they provide, that would be good.

The controller itself relies on heartbeats to and from each access point.  When enough heartbeats are missed, it marks each AP as down and the last ip address it used to contact the controller.  The "show ap database" command will show a list of up and down APs.  There are ways to obtain that information via SNMP, as well.

Thanks for the response but perhaps I wasn't clear enough.

We need WhatsUp Gold (WuG) to be able to see the AP's as UP or DOWN via as simple a protocol as ping - for that to happen we need to be able to ping the APs - at present we can't evemn ping them from the directly attached switch.

So, there must be a way (I hope) within the WLC to configure the end APs to be pingable from our monitoring station?

For the record - WhatsConnected (WC) a companion tool to WuG used for discovery and CMDB purposes amongst others can scan in the WLCs and it picks them up, their interfaces, etc and the >2K attached APs. What it doesn't allow us to directly do is monitor the up / down from this platform as WuG can not directly see the APs.

Oh.. we have our own SNMP tool that our software engineers developed years ago that just monitors systems through pings..

Colour me confused now ...

How does that work and does it allow you to directly ping the APs? It is that aspect we need to enable / get working.

---

@StuartD wrote:

Thanks for the response but perhaps I wasn't clear enough.

 

We need WhatsUp Gold (WuG) to be able to see the AP's as UP or DOWN via as simple a protocol as ping - for that to happen we need to be able to ping the APs - at present we can't evemn ping them from the directly attached switch.

 

So, there must be a way (I hope) within the WLC to configure the end APs to be pingable from our monitoring station?

 

For the record - WhatsConnected (WC) a companion tool to WuG used for discovery and CMDB purposes amongst others can scan in the WLCs and it picks them up, their interfaces, etc and the >2K attached APs. What it doesn't allow us to directly do is monitor the up / down from this platform as WuG can not directly see the APs.

 

---

What you can do is enable the SNMP trap listener on WUG :  https://docs.ipswitch.com/whatsup_gold_v12/webHelp/index.htm?enabling_the_snmp_trap_listener.htm?toc.htm

After that, you can define what traps will alert you:  https://docs.ipswitch.com/whatsup_gold_v12/webHelp/index.htm?receiving_snmp_traps.htm?toc.htm

On the controller you can configure the controller to send SNMP traps to your WUG device:  http://www.arubanetworks.com/techdocs/ArubaOS_64x_WebHelp/Web_Help_Index.htm#ArubaFrameStyles/Management_Utilities/Configuring_SNMP.htm?Highlight=snmp trap

You can then see the text of what traps are sent out like this:

(Aruba7005-US) #show snmp trap-queue

2018-01-19 08:56:14 Access point 00:24:6c:cb:30:88 with name New-AP-105 and IP address 192.168.1.230 is Down 

2018-01-19 09:06:18 Access point 00:24:6c:cb:30:88 with name New-AP-105 and IP address 192.168.1.230 is Up 

2018-01-19 09:06:18 Access point 00:24:6c:cb:30:88 with Name New-AP-105 and IP address 192.168.1.230 cold-started 1 time(s)

 This is just a stopgap measure to get things working fairly quickly.  This might not meet all ofyour needs.

Thnak you Colin. I'll take a better look on Monday but that would seem to do the job.

OK, so have had a chat with our Tools TA and we are going to go forward with this.

But one last query... do you have the spcific OID's of the AP up / down messages? Or a link to where I can download the MIBs for these devcies and we will dig out the ones we most would like to use?

We basically have the same setup you have. Over 2700 APs. We monitor the APs via pings as we do with our network switches. I've found that our monitoring system seem to be a little faster than the controller. On the flip side, there have a few times that our monitoring shows an AP down because it missed a few pings but the controller didn't only because I think that answering a ping is a low priority in Aruba. The AP came back up within a minute and nothing ever showed up in the controller..

@magnumpi83 -  That is exactly what we want to do.

Can you advise how you achieved that? I have had no involvement in the project for rolling this out, I'm just the schmuck that has to get it integrated into our monitoring setup. So I need to be able to go back to the project engineer and say "You need to do, X, Y & Z please" :)

That would be fab - thanks.

---

@magnumpi83 wrote:

We basically have the same setup you have. Over 2700 APs. We monitor the APs via pings as we do with our network switches. I've found that our monitoring system seem to be a little faster than the controller. On the flip side, there have a few times that our monitoring shows an AP down because it missed a few pings but the controller didn't only because I think that answering a ping is a low priority in Aruba. The AP came back up within a minute and nothing ever showed up in the controller..

---

The Aruba controller determines an AP is down based on missed heartbeats to the controller, not pings.  The controller requires a number of missed heartbeats before the AP is marked as down.  The controller would indicate APs that are down and missed heartbeats over the threshold in the syslog.