The current capwap delay time lower than threshold value

  • 2
  • Question
  • Updated 2 years ago
I get this message in on a classroom AP: "The current capwap delay time lower than threshold value".

The internet is kinda gone for 5 min., suddenly getting up and running again.

The other 19 AP's does not get the problem, and they are not loosing internet doing the time this specific classroom AP does.

What does "The current capwap delay time lower than threshold value" mean in a detailed explanation?
How can i perform fail searching/fix this capwap delay? Why does it appear?

:S Hope you guys can point me in some direction.
Photo of Mike Tulin Holmgaard Kristensen

Mike Tulin Holmgaard Kristensen

  • 20 Posts
  • 0 Reply Likes
  • Confused, frustrated

Posted 2 years ago

  • 2
Photo of Luke Harris

Luke Harris

  • 265 Posts
  • 18 Reply Likes
Are you using HMOL or HM On-Premise? What version of HiveOS are you running?
(Edited)
Photo of mr bee

mr bee

  • 77 Posts
  • 0 Reply Likes
Connection between AP and HM is not ok. Did you check the switchport errors ? Maybe there is a cabling problem specific for that AP.
Photo of Mike Tulin Holmgaard Kristensen

Mike Tulin Holmgaard Kristensen

  • 20 Posts
  • 0 Reply Likes
HiveManager ® Enterprise   4.1r4
Model Number AH-HM-NR-1U

Thats the hive manager information. Its OLDDDDD. But Free :D

We have the Hivemanager locally on the school. No cloud.

There has actually been some some cable errors in that building, that was never fixed for real.
I believe that just a very small disconnect is enough to show the failure?

The AP is "White" all the time, never turning "Blue" or "Green".

What should i check on even more? I did a check on switch, but didnt seem to tell me anything.
Photo of Luke Harris

Luke Harris

  • 265 Posts
  • 18 Reply Likes
In an ideal world your AP's (depending on model - can you provide this?) should be running either HiveOS 6.5r3 (gold release) or 6.5r4. Which would require an upgrade of HiveManager. 

Certainly cabling could be a contributing factor. Have you the necessary equipment to check end to end cable continuity? - checking switch ports for errors is also a worth while practice. 
Photo of Mike Tulin Holmgaard Kristensen

Mike Tulin Holmgaard Kristensen

  • 20 Posts
  • 0 Reply Likes
Thx. you very much. I will test the actually cabling and switching.
What tools can help me test the end to end cable continuity?

Any software that giefs me a good log on whats going on, if small spikes often appear?
Photo of Luke Harris

Luke Harris

  • 265 Posts
  • 18 Reply Likes
That all depends on what resources you have. I favour a Fluke Etherscope. A basic cable tester will give you a read-out but it won't be as accurate or as informative. 

Check patch cables as well. 
(Edited)
Photo of mr bee

mr bee

  • 77 Posts
  • 0 Reply Likes
check error rates on the switchports, this indicates cable problems
Photo of Steve Wood

Steve Wood

  • 5 Posts
  • 0 Reply Likes

I'm having the same issue.

Hivemanager on premise (VM), version 6.8r2. We have over 600 x AP230's running 6.5r4.

Recently noticed a few AP's will timeout connection to HM. Usually around 2 - 4 at a time, they will reconnect anywhere between 5 minutes to an hour later. Isn't always the same AP's, can be random out of the >600 AP's we have. Alarm on HM shows capwap delay time lower than threshold.

Client connectivity is affected to the AP's in question. Connection will just drop out, then re-establish again when AP is OK.

Can ping affected AP's but response time is noticeably higher than normal (20-50ms instead of 1ms). Attempting to SSH onto AP will sometimes work, otherwise times out. When connected via SSH the CLI is very slow and unresponsive. When attempting to run any command (e.g. "show cpu" or even "exit"), more often than not get an error message such as "ERROR: cli ui wait reply timeout".

Rebooting affected AP seems to resolve the issue, but cannot keep rebooting over 600 AP's!

Currently have this logged with our network support who are talking to Aerohive about the issue.

Personally I suspect a firmware problem.

Photo of Mike Tulin Holmgaard Kristensen

Mike Tulin Holmgaard Kristensen

  • 20 Posts
  • 0 Reply Likes
Are there any video guides to find, on how to use the Vprobe and log system etc. etc.
???

I could use some education.

Best regards Mike Kristensen
Photo of Luke Harris

Luke Harris

  • 265 Posts
  • 18 Reply Likes
Steve - the CPU and memory figures you quoted. Are these when the AP's are under load or sitting idle? If the latter, I would be quite concerned. 
Photo of Steve Wood

Steve Wood

  • 5 Posts
  • 0 Reply Likes

The CPU and memory figures are when the AP's are idle with no wireless clients connected.

Thanks

Photo of Luke Harris

Luke Harris

  • 265 Posts
  • 18 Reply Likes
50% CPU/Memory usage is definitely too high for an idle AP. 

I also can't see anything untoward with your configuration. Without monitoring your CPU/system processes in real-time I can't make an educated guess as to what is causing your capwap issue.
(Edited)
Photo of MST

MST

  • 152 Posts
  • 3 Reply Likes
We have similar problem since 6.6 and now with 6.8. Used PRTG and PING is ok all the time. Having 60x230s. it happens randomly , all cables have been checked with fluke a couple of times so its not the cable issue, aerohive tech advised to ignore such messages. Still troubleshoot on our own.