2/5 ROS v3.1 x86 rebooting by Kernel Panic

Hi people,

2 of 5 newest x86 PCs (Intel P4 DualCore, Core2Duo, HT) upgraded to 3.1 are rebooting by Kernel Panic, one was solved, at least for now…

The solved MT are doing Hotspot, Webproxy cache, Advanced QoS L3, DNS cache, DHCP Server, Bridging… this PC is the access concentrator for more than 20 RBs(all RB532/A and RB333 with v3.1) in mesh supporting more than 700 wifi clients.

Now this server is uptime at 21hs, i think the solution was:

/system hardware set multi-cpu=no

and

/interface bridge settings set use-ip-firewall=yes (here is a BUG… i can’t undo this option, when i try to put use-ip-firewall=no the router is rebooted without log nothing)

The other router with problem and not solved with theses settings is a simple secondary MT doing Webproxy-Cache and DNS-Cache, nothing more! is very basic but the reboot by KP is each 2-4hs, i don’t know why… tomorrow i go try to disable HyperThreading from the BIOS.

It’s any more with KP problems?

Regards!

try disabling web-proxy (or set cache size = 0), caching proxy contains bug

:open_mouth: Thanks for your reply, but the only function of this secondary router is webproxy :slight_smile:

set zero cache size, just to check whether web-proxy is the reason

or would be Max. Cache Size=none?

yessss, xzactly none =)

Without Proxy Cache, without MultiCPU, with new v3.2 the problem continue…

In logging i can’t see nothing, nothing is reported, now the problem is at 3 times at day…

Please, only in x86 is the problem, in MIPS or PowerPC all work excellent!

Any help?

I faced this problem after upgrade
and when i reinstalled ROS without keeping old configuration its worked well.