Has anyone seen this problem before.
Access points 1142 controller 5508 running code 188.8.131.52
When you set the access point port speed with "config ap duplex full speed 1000 all" the access points leave the controlle and do not rejoin. I have had to reset the the access points manually with a power down and hold the reset button.
I cannot find a bug report for this.
Whats the MAC address of the AP??
if the AP's mac-address start with YY:xx:xx:xx:xx:xx where in YY is anything other than 00?? then there is a bug.. need to upgrade the WLC to 7.0.98.
Thanks Surendra really glad it was not several hundred access points
I cant check the mac addresses till Wednesday as its a client install however I do find it quite frustrating that it appears there are more and more bugs within an assurewave release than a considerably earlier version such as 4.2. and we need to go to a newer release that is not assurewave to get around things that realistically should not be an issue.
Its also frustrating that I have raised a TAC case and explained at length the issue also about to supply the sh run-config yet I get a much quicker answer in the community about a known issue, ie bug and they could not pick up relatively quickly that its an issue.
Do you have a bug reference.
Maybe I am just tired!!!!!!
Just wondering, are your switches GIG that the APs connect to?
Am a TAC engineer.. here is the bug Id.. CSCte01087 this is the close match that am able to find.. let me know how this works out for you!!
George the switch ports are all gig poe, hard set to 1000 full. Would have been a schoolboy error but always worth checking the obvious!
Surendra unfotunately that error relates to to code 184.108.40.206, the access points join the controlle ok but if I issue the command "config ap ethernet uplex full speed 1000 all" they fail and require a hard reset.
The bug ID you suggest also appears that the aps do not join rather than fail on a specific command?
ooops..... sorry.... really sorry... the previous post was not for this.. the bug ID is CSCth98635
Just wondering why you need to hard code speed and duplex on the port? If you leave default setting (auto/auto) on both swich and AP sides, it should just end with1000/full. Than is what we do.
It could be that bug however the bug report specifically mentions autonomous aps, so what would happen is the ap is ok set with 1000 full but the switch just doesn't think its there?
The client wants speed and duplex set, also if the command it available it certainly should not break something, equally auto auto is alot better than it used to be however is far from reliable, what if the switch/ap port combination decides to set itself at 10 half using auto auto? OK it hasnt but what if it did thats obviously possible aand not good.
Its simply a case of yes there is a work around but with the client spending an awful lot of money he wants it to work how he wants it to work.
If you bought a car for £50K and you couldnt get your favourite radio station on the radio because of a bug would you be happy? No. At the end of the day I accept that there are bugs etc but saying just leave it like this is not really aceptable as we have an expectation as client that things should work as described.
I do not want to be negative but there is a lot wrong in the 6.x code that really shouldnt be and its basics like not being able to set speed and duplex, the issues with Vocera etc and saying just use auto auto is not good enough for me as an engineer. First does Csco know about it, second is there a fix, third is it something I have done.
Now lets assume this was a hospital where there were 1000 access points in ceilings and they all went off the network, that would be catastrophic. Having to find and reset 1000 access points in a hospital. I actually read the release notes for specific code trains I use and this is not in the list of bugs in 220.127.116.11
just want to make sure you also hard code the settings at switch side as well, if the switch side is left with auto/auto, it will result in half duplex.
I completely understand your concerns... n i agree for the example that you have provided.. Whats your AP model number? Does that match that of yours?? and the bug is not yet resolved.. its just opened and the development team is actively working on it.. till then they have provided the workaround.. they did not leave us in the middle of the ocean by not providing the workaround..
I request you get subscribed to the BUG details which is in the Bug link and you will get the update on that. and FYI.. if you are not happy with 6.0.196.
then 7.0.98 is already in the website from past more than 2 months...
The aps are 1142s, I do not want to start using code 18.104.22.168 yet as its stil new however have only started deploying code 6.x in the last 6-12 months as there are so many new features and the WCS is so much better but it is not as stable as code 4.2.
I guess I will have to leave it and be happy with auto config and see what the future holds hpefully 22.214.171.124 is more robust with fewer bugs
Please mark the Question as Answered, if the provided information is correct and it helped. By doing that others can take benefit as well.
Community Manager – Wireless