Network Management

We are managing a cluster of IAP-215 built by 100 nodes by AirWave. Our problem is, that the communication between the AirWave and the cluster is pretty slow. The configuration update takes sometimes up to two hours, when we run some debugging command from the GUI, it always timeouts. The status of the APs in the monitoring tab shows delayed information for more then 30 minutes. The version of the AirWave is 8.2.3.1 and it is running in a virtual environment (5 Cores, 20G RAM). We checked it status, and there is no high CPU or RAM utilization. We've also checked the utilization of some of the AP's and it looks fine as well. What could cause the problem?

Hi ,

Both IAPs and Airwave server is in same location or managing the IAPs devices remotely? If we are managing rmeotely make sure you have good bandwidth.

 

How many devices are you montiorng ?

 

As you mentioned that sometime IAPs command runnnig from Airwave is timing out, you could try increase the https timeout value in Groups>Basic>Aruba Instant section page and check the status. Airwave communicate with IAPs using https.

 

Regards,

Pavan

If my post address your query, give kudos:)

 

 

Hi Pavan, the AirWave manages 720 IAP-215 APs. 100 of these devices are local devices separated from the AirWave by firewall, the rest of the devices are remote connected by an MPLS link. The problem was noticed with the local AP's but it is possible it occurs at the remotes as well. I've checked the HTTPS Timeout Time, and it is set to 5 minutes. I think it is not too short.

Could you share complete system>performance page screen shots.

Regards,

Pavan

Sure. The screenshots are in the attached zip file.

Server performance is pretty good, it might be issue with your network. We could check whether any latency in your network by ping. Try ping the IAPs from Airwave CLI and see the response time.

 

You could install qperf tool on Airwave and IAP (not sure exactly we could install on IAP side) to test the bandwidth and latency.

 

https://access.redhat.com/solutions/2122681

 

Regards,

Pavan

 

Hi Pavan, we've made a ping test from AirWave to two APs in the cluster (master .10, member .135) and it looks fine. Here is the result.

[root@mercury]# ping -s 512 -c 30 10.95.230.10
PING 10.95.230.10 (10.95.230.10) 512(540) bytes of data.
--- 10.95.230.10 ping statistics ---
30 packets transmitted, 30 received, 0% packet loss, time 29016ms
rtt min/avg/max/mdev = 1.241/1.712/9.439/1.565 ms

[root@mercury]# ping -s 512 -c 30 10.95.230.135
PING 10.95.230.135 (10.95.230.135) 512(540) bytes of data.
--- 10.95.230.135 ping statistics ---
30 packets transmitted, 30 received, 0% packet loss, time 29006ms
rtt min/avg/max/mdev = 0.891/1.264/1.631/0.109 ms

[root@mercury]# ping -s 1024 -c 30 10.95.230.10
PING 10.95.230.10 (10.95.230.10) 1024(1052) bytes of data.
--- 10.95.230.10 ping statistics ---
30 packets transmitted, 30 received, 0% packet loss, time 29022ms
rtt min/avg/max/mdev = 1.499/1.590/2.387/0.171 ms

[root@mercury]# ping -s 1024 -c 30 10.95.230.135
PING 10.95.230.135 (10.95.230.135) 1024(1052) bytes of data.
--- 10.95.230.135 ping statistics ---
30 packets transmitted, 30 received, 0% packet loss, time 28997ms
rtt min/avg/max/mdev = 1.318/1.600/2.837/0.246 ms

[root@SK3-AirWave mercurping -s 2048 -c 30 10.95.230.10
PING 10.95.230.10 (10.95.230.10) 2048(2076) bytes of data.
--- 10.95.230.10 ping statistics ---
30 packets transmitted, 30 received, 0% packet loss, time 29013ms
rtt min/avg/max/mdev = 1.750/2.056/3.598/0.366 ms

[root@mercury]# ping -s 2048 -c 30 10.95.230.135
PING 10.95.230.135 (10.95.230.135) 2048(2076) bytes of data.
--- 10.95.230.135 ping statistics ---
30 packets transmitted, 30 received, 0% packet loss, time 29002ms
rtt min/avg/max/mdev = 1.900/1.954/2.271/0.083 ms

We're working on the nperf testing, but it will take some time.

As you mentioned there is a firewall in between Airwave and IAPs, could you allow jumbo frames in your firewall. AMON packet size will be large and if we dont allow jumbo frames their might be chance of packet drop.

 

Regards,

Pavan

Hi Pavan,

we checked the jumbo-frames settings. Jumbo frames are not enabled and there is no way to enable it :-(

We need to allow jumbo frames in firewall and also try add test IAP to Airwave where there is no firewall in between and see the communication delay.

 

Regards,

Pavan

Hi Pavan,

 

just to know. We're working on how to enable jumbo frames in the infrastructure, but it looks it will take quite a long time.

Hi,

 

We need to allow jumbo frames in your firewall, if you are using AMON.

 

If you are seeing huge dealy in data update, I would recommand to open TAC ticket, so that engineer could get chance to check the server settings.

 

Regards,

Pavan