Hardware

Hi

I'm using m0n0 wall since ~3 years on a wrap board and since a few months I'm experiencing many issues with the stability. It normally runs fine, but sometimes (about once a week) it just keeps crashing every 5 minutes for hours. Since one of the later versions (I'm using 1.3b4 as of now, but the latest stable is even worse) it automatically resets itself, so it is not much of a problem, but it's very annoying to see all connections go down all the time.
Did anyone experience similar problems using one of the wrap boards? Or has anyone any idea, what could be causing these problems?

I can't seem to use m0n0wall or pfsense at all on these boards and soekris because it will kill all wan connections under heavy traffic. Seems to me like freebsd is the culprit.

So do you use m0n0wall on another system? On what system, if I may ask? And do you have any idea, why the problems only appeared in the last few months (it went well for about 2 years)?

What other systems can you recommend? It should be very low power and have passive cooling, thats why i chose the wrap boards in the first place.

Just to clarify: I don't use any p2p tools on a regular basis (only torrent for some larger downloads) and it also happens when having only a few (5-10) open TCP connections.

we use many wrap1e203 on customer sites running m0n0wall (1.2x). Because your board was running fine for a long time and suddenly starts acting strangely, one can't ignore the option of having a hardware problem, ranging from a faulty power supply to the on-board memory losing occasionally some bits.

Do you have access to another power supply and another wrap board to verify?

I'll test exchanging the power supply, but the board may be a bit of a problem. I'll see, if I can convince someone to swap it .

Hi,

Have you tried changing PSU's? I am experiencing the same. It started out with the DHCP crashing about once a week. Now, I'm happy if m0n0 keeps running for a day.

I'm using a wrap board. It has been running absolutely fine for more than a year, no issues whatsoever.

Greetings from a cold Netherlands...

I've got news.

1. Exchanging the PSU did not help at all.
2. It does not seem to be (solely) a hardware problem. Updating to the latest beta caused my monowall to crash every 5-30 minutes.
3. I now have a console output of a crash, but it does not look very good:

Code:

Fatal trap 9: general protection fault while in kernel mode
instruction pointer     = 0x20:0xc0672929
stack pointer           = 0x28:0xc8795d00
frame pointer           = 0x28:0xc8795d00
code segment            = base 0x0, limit 0xfffff, type 0x1b
                        = DPL 0, pres 1, def32 1, gran 1
processor eflags        = interrupt enabled, resume, IOPL = 0
current process         = 10 (idle)
trap number             = 9
panic: general protection fault
Uptime: 26m3s
Cannot dump. No dump device defined.

Is there any way to get a dump? Mounting the CF read write? Network dump?

Any help / idea is appreciated.

I'll try downgrading to some older versions to see, if any of them is better. Maybe I'll find one that works without any problems. I'll also try out pfSense as soon, as I get a larger CF card, to see if it's a monowall problem.

I agree that it's not solely a hardware problem because of the following. My WRAP board crashed twice a day in the end. As it's running a production firewall (for 8 users in a small company), these users got very annoyed.
So, I swapped the PSU by one I ordered from pcengines in the end I even swapped the WRAP for an ALIX...

I tried this:
1) Replacing the PSU --> didn't solve stability problems
2) Replacing the WRAP by an ALIX --> a lot more stable but still crashing once a week approx.

But, the weirdest thing is, I used the exact same board for a couple of weeks in a heavily used test environment to make sure it was stable before I moved it to production...

So, I think it's either:
1) a corrupt configuration item somewhere in m0n0
2) a corrupt flash-card which gives read-errors once in a while and the ALIX board handles these better than the WRAP board

I came to this two option, because I used the production flash in the new board and it crashed 8 days later and a seperate card for the test env which didn't crash for weeks on end.

I will test the last hypothesis this weekend and post updates when it has or hasn't crashed after a couple of weeks.