Losing access to network

[ISoliman]

What i know is that the 5.5.3.0 branch had a big issue with clines disconnecting that's why I recommended using another branch like 5.4.x, the only thing that forces you to use the 5.5.3.0 is if you have the new APs, other than that i won't recommend using it, I received a new version from HP that solved the disconnection issue (might be the same issue you are facing) but to be honest it might be with the interference you have since you mentioned that the APs are close to each other and yes this will cause the clients to disconnect also, so 1st of all use 5.4.x if you don't have any of the new APs 460 or 430, then decrease the power automatically in all the APs to avoid the interference (better to do a site survey), also there is an option under the Radio configuration for the APs where you set the distance between the APs, to know more about this option please check the management and configuration guide.

[Craig-Lyndes]

Again, Thank you ISoliman for your time.

I switched to the 5.5.3.0 because I added 3 MSM430 radios (replaced MSM410's in classrooms that were having an issue) to accomodate our 1:1 initiative where every student is given a laptop.  60+ students connecting simultaneously to a MSM410 locked at 802.11G was not pretty.  That is also why my access points are relatively close together.  I have a situation where I can have a flash mob of up to 60 devices at any time in any classroom (we have teams of 3 teachers with roughly 20 students each, so when they get the team together for a group activity is can be quit a challenge for the wireless system). 

I have the radios set for Distance between APs: Small and I have the auto power turned on as you noted.

I'm still suspicious that the Rx FCS errors and the disconection from the network are 2 different problems.  One point I made earlier is that having a Mac with an Apple Airpoirt wifi chip set come into range of the access point seems to be able to trigger the loss of network - sometimes.  This has happened enough so that I believe it is more than a coincidence.  This seems to go along with the solution of locking the access points to 802.11G, it probably has to do with guard frames and the Mac jamming.  I'm also suspicious that the problem only occurrs when there is a lot of network traffic.  Again probably something to do with timing.  I'm not a RF engineer though, so some of this stuff isn't very clear to me.

Thanks again for looking at this!

Craig Lyndes

St Albans City School

[ISoliman]

Ok then since you have 430 APs then ask HP for the latest maintenance release for the 5.5.3.0 (not the one you have there is a new one).

Also for 5.5.x it has a new feature called "Band Steering", you can use that to allow 1 430 AP to support the maximum number of clients if the clients are having dual ban wireless cards, this will allow the controller to shift the clients that can work in 5 GHz to connect to Radio 1 so that both Radios can be utilized, in this manner you can support between 50-60 users per AP.

For the Airport wifi chip you can simply reproduce this by getting a mac machine and trying to access the wireless on specific area and see if it will show the same result, if you are able to replicate it 100% everytime you use the mac machine then simply capture the wireless traffic from the AP side (from the controller go to groups then click on the AP then on tools and choose wireless trace/capture and start it, then replicate the issue and collect the logs and the capture and log a case with HP stating all the steps you have done so that they can reproduce the issue from their side also this is the only way to get this fixed if it is a bug :).

FYI, I have a setup with 460 APs and clients where disconnect a lot same as what you have mentioned, I used the latest version provided by HP and after that only everything worked fine for me :), I remember that the issue was showing even when no Mac machines are nearby, might be that some users where using the iphones not sure about that but anyway it was solved and i'm glad :) hope that latest version will work with you also

[tschaps]

I was experiencing a similar situation with my MSM430's in autonomous mode before we got the controller working. It would be interesting to find out if the little trick I discovered works in your situation: when you have a laptop that has lost connectivity (but still has IP and a wifi signal) ping the gateway. If like mine, it should time out/fail. Note its IP address. Then ping this laptop from ANOTHER computer on the same subnet, it should work, and then see if all of a sudden your connectivity is restored. 

Note, I don't tout this as a solution, but if it is the same behavior, it may help isolate the problem. 

[BGraham_1]

I believe HP finally has the problem fixed. We have been giving them information and working closely with the support rep over the past few months. THey sent us an engineering release they sent us a couple of weeks ago and I installed it. Since then, we have not been getting any 169.254.x.x addresses (we were only getting these on Radio 2) and we have not been getting any IP addresses assigned to the clients from the APs untagged (management) VLAN on the MSM460s.

They did not say when they are going to release it. I asked if I could share it, but they said no.

You might want to open up a case with support and ask for version 5.5.3.0-fa2-10564.

I hope this helps.

Bob

[temuri426]

hello, 

  Every second day users are losing conectivity.  after restarting devices several times AP's are working ok.

   there are some messages in logs , can ou please explain  what does this means: 

Dec 19 21:41:47 warning maestro_mas SG1383N2R6 Cannot take proper handling action on event from the system message bus, skipping (event='evLocalFirmwareRetrieval')

Dec 19 21:43:33 warning maestro_mas SG1383N2R6 Last message repeated 18 times

Dec 21 16:26:23 err kernel SG1383N2R6 e1000: eth1: e1000_clean_tx_irq: Detected Tx Unit Hang

Dec 21 16:26:40 err kernel SG1383N2R6 Last message repeated 4 times

Dec 21 16:26:40 warning statspoller SG1383N2R6 [44:1E:A1:C2:80:E2]: Transfer of SC statistics information failed: (28/0)

Dec 21 16:54:09 warning maestro_mas SG1383N2R6 Cannot take proper handling action on event from the system message bus, skipping (event='evLocalConfigChanged')

Dec 21 16:57:53 warning mapconf SG1383N2R6 SOAP configuration process <29215>: terminating upon maestro request.

Dec 21 16:59:35 warning maestro_mas SG1383N2R6 Cannot take proper handling action on event from the SC management message bus, skipping (event='evLocalTunnelDown')

Dec 21 17:00:54 err maestro_sla SG1383N2QV Failed to identify discovery port.

Dec 21 17:04:24 warning webs SG1383N2R6 Logout requested by login takeover

Dec 21 17:12:23 warning maestro_mas SG1383N2R6 Cannot take proper handling action on event from the system message bus, skipping (event='evLocalConfigChanged')

Dec 21 17:14:49 err mapconf SG1383N2R6 SOAP FAULT: SOAP-ENV:Client "Invalid configuration"

Dec 21 17:14:49 err mapconf SG1383N2R6 Detail: <error><errorcode>1009</errorcode><errorinfo>Configuration commit timeout.</errorinfo></error>

Dec 21 17:14:49 err mapconf SG1383N2R6 [98:4B:E1:E8:23:9E] CONFIG_full_vsc_security() failed, VSC <Agruni/2/2>.

Dec 21 17:14:58 warning maestro_mas SG1383N2R6 Cannot take proper handling action on event from the SC management message bus, skipping (event='evLocalTunnelDown')

Dec 21 17:16:30 warning monitord SG1383N2QV Watchdog Update after 32 seconds uptime=156855

Dec 21 17:16:30 err maestro_sla SG1383N2QV Failed to identify discovery port.

many thanks

[mikeguk]

We have been having the same issue with our 760.  

In case others are interested, 5.7.0 is now out and lists the following as an explicit fix

https://h10145.www1.hp.com/Downloads/SoftwareReleases.aspx?ProductNumber=J9421A&lang=en,en&cc=us,us&prodSeriesId=3963981

"47397 Devices become unreachable on the management LAN after 20 to 30 minutes."

Installed this morning and now waiting to see if the problem is finally resolved.

[temuri426]

hi all, 

 after updating to 5.7.0 , it looks like the problem is fixed . 

p.s when we've upgraded firmware problem repeated once again after two dayes, but i restarted all devices and now everything works good 

                   12 days uptime :))  cool

[goAtsy]

mikeguk that fix is in the release notes for Firmware 5.4.0. HP must have posted the wrong document. This page links to the correct release notes for 5.7...

https://h10145.www1.hp.com/downloads/SoftwareReleases.aspx?ProductNumber=J9325A

[nogirs]

Hi eveybody, I read through this interesting thread and as we are having the same trouble. I've gotten confused as to the the many solutions mentioned.

Is 5.7 the answer or has 5.3 provided the best possiblity stable solution?

Also, we're on 5.4.2 - can I jump to 5.7 or should I do the ladder (5.3, 5.5, 5.7)?

Thanks