6.1.2 LSVPN/VPN Hang - Reboot Required

[dusk2dusk]

I am having some issues as well with 6.1.2 and LSVPN to hub which is on 6.0.5h3.  This is vaguely referenced here by others Anyone use 6.1.2? Is it stable where everyone affected, says yeah there's this IPSEC issue and never actually describes what they are seeing.  So I'm going to detail what I am seeing.  

I had two PA-200 devices overnight go offline, same ISP(Could have been coincidence) on US-EST but at different times.  My routers could see each other properly at either end but the PA-200 satellite devices would not respond on their public facing IP's at all.  I could route right up to the devices but could not get in on https via port 4443 or via ssh.  Definitely seemed like a bug that might have been caused by some blip in the ISP maybe but the only way to fix it was to reboot the devices.  As soon as they rebooted LSVPN came back up.

Reading through the satellite's syslogs at the time the device went down, there's no particular indication other than 'GlobalProtect Satellite connection to gateway failed.  Satellite failed to connect to Gateway 'IP-w.x.y.z' due to "connection failed".'  This error continued until we rebooted the PA-200's at which point they immediately came back.  Devices showed in ARP on routers and everything but they wouldn't respond whatsoever to any Layer3 requests/comm.  It's like the security and interface management policies just plain failed. 

*** An additional but somewhat related issue I have been having is with devices indicating active LSVPN connections but in fact they are not passing any traffic.  Tunnel monitor is set to the hub device's tunnel interface IP.  This only happens to one device at a time and all other of my 40 or so PA-200's stay active without issue.  It seems sometimes this happens when the IP changes in a failover scenario and others it seems like it just drops off after a rekey.  I know there are some fixes in indicated in 6.1.2 for this and since my hubs are not yet on 6.1.2 I did not reach out to support on this yet.  But now I don't think we can go to 6.1.2 if there is a known bug/issue.  I had to have people drive in and reboot these thigns in the middle of the night.  Needless to say they are not feeling very happy about our 'upgrade' to Palo Alto from tried and true but old Netscreen/SSGs.

[LCMember2041]

This is bug in panos (any version). My problem is similar but effect is the same – FW hangs and reboot required.

Palo write:

·         > Do we know what is the main reason of this problem ?

·          

·         so far we know that one of the internal timers gets stuck, it is not

·         clear in which conditions that happens

·          

·         > Is there any configuration change we can do to minimize these downsides ?

·          

·         disabling the qos, which you already did is the only one known so far to

·         limit how often it happens

·          

·         > When do you plan to fully address this issue ?

·          

·         once the releases with additional debugs are out in the field this

·         should accelerate the resolution. No other time frame can be provided at

·         the moment

·