RV220W rebooting at random intervals

[NeonNero1]

Hi,

Over the past few weeks, our RV220W router has been experiencing random reboots throughout the day, causing a network outage for our local network for the few minutes it takes for the router to start up again. We've seen it reboot at least 5 today alone (and the time is currently just a little over 1pm).

I'm having trouble finding the reason for these reboots in the router's logs, although I find these two lines repeating in the error log:

Mon Jan 28 13:20:52 2013(GMT+0100) [rv220w][System][EVTDSPTCH] umiIoctl (18, UMI_CMD_DB_UPDATE(4)) failed. table=dot11STA row=11
Mon Jan 28 13:20:52 2013(GMT+0100) [rv220w][System][PLATFORM] /pfrm2.0/bin/bwLimitConfig /tmp/system.db 18 dot11STA 11 failed. status=-1

I'm not sure why these messages pop up, since we're not using the bandwidth limiter module of the router (granted, we do use QoS profiles to ensure our own LAN is prioritized over the guest VLAN). I'm also not sure whether this is related to the router reboots.

I also find the following repeated line in the warning log:

Mon Jan 28 12:50:50 2013(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK

The router is currently running on firmware version 1.0.4.17 (latest official release), since we needed support for hairpin connections (as I found out in Public IP used internally on RV220W). I've seen the suggestion of downgrading to a previous firmware version in RV220W random reboots, which has possibly solved it for the poster in that thread, but as we really need the hairpinning support, I don't know which firmware version is safe enough to downgrade it to, if needed.

Any suggestions, either for a solution or what else to look for?

[NeonNero1]

This problem still exists, so here's a follow-up.

This is the entire warning log (default), after it rebooted again about 10 minutes ago:

Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] CVMSEG size: 2 cache lines (256 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] CPU revision is: 000d0601
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Determined physical RAM map:
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL]  memory: 0000000007800000 @ 0000000000700000 (usable)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Built 1 zonelists.  Total pages: 30276
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Primary instruction cache 32kB, virtually tagged, 4 way, 64 sets, linesize 128 bytes.
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Primary data cache 16kB, 64-way, 2 sets, linesize 128 bytes.
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] PID hash table entries: 512 (order: 9, 4096 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Using 400.000 MHz high precision timer.
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Dentry cache hash table entries: 16384 (order: 5, 131072 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Inode-cache hash table entries: 8192 (order: 4, 65536 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Calibrating delay using timer specific routine.. 800.39 BogoMIPS (lpj=1600784)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Mount-cache hash table entries: 256
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Checking for the multiply/shift bug... no.
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Checking for the daddi bug... no.
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Checking for the daddiu bug... no.
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Enabling Octeon big bar support
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] PCI Status: PCI 32-bit
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] IP route cache hash table entries: 1024 (order: 1, 8192 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] TCP established hash table entries: 4096 (order: 4, 65536 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] TCP bind hash table entries: 4096 (order: 3, 32768 bytes)
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] /proc/octeon_perf: Octeon performace counter interface loaded
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Total HugeTLB memory allocated, 0
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] RAMDISK driver initialized: 1 RAM disks of 500000K size 1024 blocksize
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] GACT probability NOT on
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Mirror/redirect action on
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] u32 classifier
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL]     Performance counters on
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL]     Actions configured 
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] nf_conntrack version 0.5.0 (480 buckets, 3840 max)
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] ip_tables: (C) 2000-2006 Netfilter Core Team
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] netfilter PSD loaded - (c) astaro AG
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] ipt_time loading
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] ip6_tables: (C) 2000-2006 Netfilter Core Team
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] ip6t_time loading
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Bootbus flash: Setting flash for 32MB flash at 0x1dc00000
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL]  Amd/Fujitsu Extended Query Table at 0x0040
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] phys_mapped_flash: CFI does not contain boot bank location. Assuming top.
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] BOOT BUS: Address decode error
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] cavium-ethernet: Cavium Networks Octeon SDK version 1.7.3, build 264
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Register the sysctl for hardware offload succeed 0 
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Register the sysctl for flow cache succeed 
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Register the sysctl for brcm Tag succeed 
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Register the sysctl for vlan enabled succeed 
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Pool 0 size 2048 No 1024
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Pool 1 size 128 No 1024
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Pool 3 size 128 No 10240
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] global cache allocation start 0
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Ipforward global cache allocation sucessfull
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Registerd the sysctl for IMP port status
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Registerd the sysctl for 802.1x enabled port
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Registerd the sysctl for blocking non EAP packets status
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Interface 0 has 3 ports (RGMII)
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 999
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 1002
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 1040
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 1042
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] port ko 0
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] port ko 1
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] port ko 2
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] led module init...
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] umi: module license 'unspecified' taints kernel.
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] eth0: 1000 Mbps Full duplex, port  0, queue  0
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] eth1: 1000 Mbps Full duplex, port  1, queue  1
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] PF_key: Initialized Security Policy Database.
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] IPsec: Initialized Security Association Processing.
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] BUG: at kernel/softirq.c:138 local_bh_enable()
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Call Trace:[][][][][][][][][][][][
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Intializing the 8021P-DSCP Remark Module
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Registerd the sysctl for _8021p task IMP port<6>registering backend engine with xlr8Nat framework
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Registerd the sysctl reboot status
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] PCI: Enabling device 0000:00:03.0 (0000 -> 0002)
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] vap0: Broadcom BCM432b 802.11 Wireless Controller 5.60.120.9 
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:52 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:52 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] HTB: quantum of class 10001 is big. Consider r2q change.
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] HTB: quantum of class 10002 is big. Consider r2q change.
Sat Jan  1 01:01:53 2011(GMT+0100) [rv220w][Kernel][KERNEL] HTB: quantum of class 10003 is big. Consider r2q change.
Sat Jan  1 01:02:43 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Sat Jan  1 01:02:43 2011(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Tue Feb  5 12:23:16 2013(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK
Tue Feb  5 12:23:16 2013(GMT+0100) [rv220w][Kernel][KERNEL] ERROR:endChantRecv: protocol PNAC returned !OK

And this is the entire notification log:

Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Linux version 2.6.21.7-Cavium-Octeon (root@localhost) (gcc version 4.1.2 (Cavium Networks Version: 1_6_0, build 34)) #1 Thu Jul 12 12:03:16 IST 2012
Sat Jan  1 01:01:50 2011(GMT+0100) [rv220w][Kernel][KERNEL] Kernel command line:  bootoctlinux 0x2a00200 mtdparts=phys_mapped_flash:512k(bootloader)ro,6M(kernel),16M(rootfs),1024k(data),128k(bootload-env) console=ttyS0,115200
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] slram: not enough parameters.
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] number of CFI chips: 1
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] cfi_cmdset_0002: Disabling erase-suspend-program due to code brokenness.
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 5 cmdlinepart partitions found on MTD device phys_mapped_flash
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] Creating 5 MTD partitions on "phys_mapped_flash":
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 0x00000000-0x00080000 : "bootloader"
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 0x00080000-0x00680000 : "kernel"
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 0x00680000-0x01680000 : "rootfs"
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 0x01680000-0x01780000 : "data"
Sat Jan  1 01:01:51 2011(GMT+0100) [rv220w][Kernel][KERNEL] 0x01780000-0x017a0000 : "bootload-env"

The other logs on the upper levels (emergency, alert, critical and error) are all empty.

Apparently, where it starts saying "Tue Feb 5" is where the date/time is updated from the NTP service, so the clock is also reset on reboot for some reason.

Any further ideas what might be the cause, or where to look next?

[larsthestar]

I have this problem as well and 3 of my RV220W routers are rebooting at random intervals. If you Google this problem you will see a lot of other people are experiencing the same. I can't understand why such a horrible firmware was ever released. I will end up downgrading to the previous firmware, that was at least more stable even though I will miss some features such as between vlan firewall and improved vlan management.

Been looking at replacing them all together and been checking out Netgear, TP-Link and DLink. So far difficult to find one that covers all your needs.

[NeonNero1]

The closest answer I've found in these forums, though, is to either downgrade to a previous firmware version, or reset the configuration to factory default and re-enter the configuration manually (which, for some reason, is recommended course of action anyway after a firmware upgrade on some Cisco routers, judging by some of the responses to similar issues in this forum). Neither option is appealing to me, and the latter option also means that the local network (along with the Exchange and VPN server placed on the local network) will be unavailable while I'm reconfiguring, and all wireless network clients that have connected to the router might have to be reconfigured aftwards as well.

On the other hand, while sending out a notice to our company staff about the possibility of reconfiguring the router (the second option I mentioned), I got a reply from my boss that he might have a second router of the same model that I can pre-configure and quickly replace the router on Monday morning, and by doing so, minimize the downtime required for the reconfiguration.

[larsthestar]

I have tried to wipe out the router completely and start from scratch again but to no avail. I still get the same random reboots. I even tried to run it with as less configuration as possible to see if my configuration broke things but still it keeps rebooting.

I thought maybe it was a bad piece of hardware, but no, as I mentioned I have two more of these and they do the same. So I think you will end up downgrading the firmware or buy a different router.

[jonatrod]

Good morning

Hi  , thanks for using our forum, my name is Johnnatan and I am part of the Small business Support community. I have seen this problem before, you could go Firewall> Basic settings and disable  “Block Multicast Packets, ICSA settins, and attack prevention option,  and the amount of random reboots will decrease . You could disable options that you don´t need in these parts and the reboot is going to decrease, I encourage you to take a look of this post https://supportforums.cisco.com/message/3498114#3498114 .

Also another solution  is downgrade to firmware v.1.0.3.5,

I hope you find this answer useful,

*Please mark the question as Answered or rate it so other users can benefit from it"

Greetings,

Johnnatan Rodriguez Miranda.

Cisco Network Support Engineer.

[larsthestar]

Can I just ask if this is now a known bug and will be addressed in future firmware upgrades?

As for me, I ended up downgrading to 1.0.1.0 as the VLAN administration in 1.0.3.5 is very strange to me with only (only Tagged and Untagged and no Exclude option).

[NeonNero1]

Thanks for replying.

Yes, I already saw that (or the same advice in a similar thread, I don't remember). I've disabled all of the options on that page, but it still reboot several times during the day, and it disrupts the workflow at the office, since our work depends on our debugging environment being able to reach both a database server at an off-site datacenter (for both configuration and caching) and data from other websites.

Since our boss told me he had bought another identical routers when he bought this one, it will allow be to install downgraded firmware and reconfigure the router from scratch without too much downtime (only when switching out the existing router with the reconfigured one).

Then again, does the 1.0.3.5 firmware include support for hairpin connections, ie. using the router's public IP address to connect to a NAT'ed server?

[jonatrod]

Hi Kim, in order to help you, I have a couple of questions for you.

Could you describe me a little more your topology?

What do  you have connected to the router.( switch, phone, UC etc.)?

What Internet connection type ?

Are you able to reproduce on demand?

In order to provide further assistance you could send me your config file to my mail (jonatrod@cisco.com )

However if you need a quicker answer, you could contact our support team here.

I hope you find this answer useful,

*Please mark the question as Answered or rate it so other users can benefit from it"

Greetings,

Johnnatan Rodriguez Miranda.

Cisco Network Support Engineer.

[NeonNero1]

Hi Jonathan,

The Internet connection is VDSL, and the RV220W router connects to the DSL router-modem via a switch in our server room. The first LAN port on the RV220W router is connected to a switch that serves our internal LAN, and this is assigned as VLAN 1. To that LAN switch, there are two servers (Windows SBS 2011 and Windows 2008 R2) and several LAN clients (Windows and Mac computers).

We also have two wireless networks set up on the router; our corporate network (assigned to VLAN 1), and our guest network (assigned to VLAN 2). VLAN 1 is set up with DHCP Relay routed to the Windows SBS 2011 server, and VLAN 2 is set up to use the internal DHCP server on the router itself. QoS is set up to prioritize VLAN 1 first.

The Windows SBS 2011 server also set up to serve our VPN (and this configuration works well, except during one of these router reboots, of course).

The wireless network serves mostly mobile devices (smartphones and tablets) and a few laptops, but I'm guessing this isn't relevant at the moment.

I'll send you the config file in a separate e-mail.

Right now, I'm leaning towards setting up the second router with the same configuration (copied manually to my notes) and a downgraded firmware version.

Regards,

Kim

[jonatrod · Accepted solution]

HI Kim, what I already check your configuration, I had no problem with the device, however I could recommend you that try to configure just the necessary settings, it looks like a little overload, you can also get another RV220 and work with two devices, in this way you are going to get a better performance. Please let me know know of any progress when you downgrade your firmware. I apologize for the inconvenience

I hope you find this answer useful,

*Please mark the question as Answered or rate it so other users can benefit from it"

Greetings,

Johnnatan Rodriguez Miranda.

Cisco Network Support Engineer.

[Bas van Oerle]

Hi,

I got exactly the same problems. Switched between firmwares but it doesn't matter. Also changed the wireless vlan to other vlan but still the same issues. Router is rebooting couple of times a day. Allready uncheck all the settings under the attack prevention. When I disable wireless it is going much better and the errors disappear.

I am so not happy with this device. I hope there will be a firmware update very very very soon!!!

[Catalin Burla]

Hi.

I have two RV220W in use for more than one year and I can confirm that under heavy traffic unit  reboots - in my case even 2--3 times per day. Turning off different features does help, but when traffic excedes router capability it will reboot.

I had 100 Mbit internet connection (guaranteed) and other routers to network segments with a total of 200-300 users behind RV220W (in router mode, by the way, wireless off, etc) which generates more than one-three thousand (peek) connections simultaneous.

After changing network topology , moving traffic away to other equipments , random reboots dropped to .... 1-2 or less  / month (still having 100 Mbps, but less clients).

Best regards,

Catalin

[NeonNero1]

I just realized that I hadn't marked this thread as answered (judging by the two new replies I was notified about), so I'll update my status on this issue.

What I ended up doing, was wiping the configuration, installing the 1.0.3.5 firmware, and setting up all the LAN and firewall settings by hand (ie. not restoring from backup). I currently have all features enabled under Firewall -> Attack Prevention, except "Block Multicast Packets" (which is left disabled). Pretty much else (aside from firewall access and forwarding rules) is left at their default settings.

The most important fact is that during the firmware install, I also set the option to wipe the configuration (based on advice from another forum thread), and manually re-entering the firewall rules and forwarding settings afterwards.

This was set up back in mid-February, and we haven't had a problem with it since.

[Dan Miley]

Thank you for the update.