Smartzone 100 and R730 Reboots/ Offline Issue

[kevin_7278725]

Is anyone having issues with R730's randomly going Offline and requiring the AP to be rebooted in order to get it back online?  We have been seeing this issue for about 30 days now and support has been unable to tell us why the R730's are doing this.  The AP's are inaccessible most of the time when the issue happens, but sometimes we are able to still ping the AP and actually SSH to the login prompt, but the admin account/password will fail.  Once we power cycle the AP it works fine again.   When the issue is happening the AP is still accepting clients but it has no network access so those clients are broken.  It is very frustrating and Ruckus support has been no help.  The R730's are connected to Ruckus ICX 7650's via 5Gb multigig ports.  The switches report no problems and there are other AP's on the same switch at the time that have no issue, so the problem is just random Access Point specific.  The issue is completely random, no pattern can be found, other than support telling us they are seeing AP kernel panics and that they can't tell us why or how to make stop.   

Smartzone 100 version is 5.1.2.0.302 - which support had us upgrade to as they said that would fix the kernel panics - It has not

R730 version -  5.1.2.0.373

A few of the R730's have not been able to recover from this issue after a reboot and have had to be RMA'd.  Some of them will automatically reboot after 15-30 mins, but if we manually reboot them they typically come back online and work.  Was curious if any else is experiencing this issue with R730's, Smartzone 100's, and ICX 7650's?  

[kevin_7278725]

Adding an update that we are still having problems.  The AP Firmware 5.1.2.0.1013 they provided us seems to have fixed one of the AP kernel panic issues.  We still have at least 2 other AP problems/bugs which are only happening a few times  per day, which is helpful, but still horrible that we randomly having R730's go offline.  We discovered another issue Friday which Ruckus support/engineering believes is a problem with their current Smartzone code that is causing their Cassandra database to error and have problem with the control plane responding properly to the AP heartbeats.  Still providing logs and packet captures daily to support with no expectations for when they will fix all these bugs and provide us with a stable solution.  Extremely frustrating and disappointing that Ruckus is not providing a stable WiFi solution and we are having to spend countless hours troubleshooting their product.

[michael_brado]

Hi Kevin, what's your current open case number please?I want to have a look (from the inside), to see what I can see for you.

[kevin_7278725]

Noted above but it is 965817

[michael_brado]

Thanks, bug ID = ER-7736, and I see Eng has a test bed and is reviewing logs collected.
Still active... did they ask about running a Debug image (that collects extra stuff) on at least a couple APs?

[kevin_7278725]

They did and that was refused by my management as you all have no way of being able to predict which AP will have the issue. Support could also not explain the impact of this debug image on our AP's. All they could say was that they don't recommend putting it on all the AP's, and only recommend a few, which makes no sense as it is completely random as to which AP's have the issue and without us knowing the impact as well as being told the image has not been tested by your engineering team made my management refuse to be your testers.  This image could possibly make our wifi even worse then it currently is performing.  Every AP, at some point, over the past 6 weeks has had one or more of the 4 issues we have discovered, with no current pattern detected, so putting an unknown image onto a few AP's seemed pointless.  The odds of guessing the correct AP was not worth the risk of running an unknown image.