the Everex m/c with PIAF latest build just went down abruptly

[sims]

any way of diagnosing what may have happened? my house was without the usual phone connectivity for a few hours (and i was out ).

wondering if i can diagnose what happened?

----

some pointers. when i was back, and i tried to ssh into the server, it said server is down. when i tried to ping it, the message was that "host is down". once it is up, there is no response to ping but not "host is down" message either.

of course the web interface was not coming up either.

and i know for sure that the inbound trunks were not registered because the calls were going to the fallback number.

 

[wardmundy]

Is the server on a UPS with a reliable battery? Any clocks blinking in the house?

 

[jroper]

Hi

The PBX in standard configuration will not respond to pings once PiaF is installed (it will during the install process)

The security reason for this is that you can work out the OS from the TTL, and if you know the OS, then you have a chance of working out what the vunerablilities of that OS may be, and therefore the best way of attacking it.

You would need to put a monitor on the PBX to see what happens when you boot.

Joe

 

[sims]

it is not on a UPS. and no, there are no clocks blinking. if there is a power failure, then the cooking range clock resets.

it seemed that at the very least the OS closed the network connection (and since there was no console connected, I have no idea what happened).

 

[sims]

that is fair. after i restarted the box, ping did not respond, but i could tell that ping response is suppressed.

before i restarted the box, the message was that the "host is down"".

no console attached to the box and i do not even have a spare one...

Hi

The PBX in standard configuration will not respond to pings once PiaF is installed (it will during the install process)

The security reason for this is that you can work out the OS from the TTL, and if you know the OS, then you have a chance of working out what the vunerablilities of that OS may be, and therefore the best way of attacking it.

You would need to put a monitor on the PBX to see what happens when you boot.

Joe

 

[bigdog101]

Do u have your IP addresses fixed on your local netowrk? Are you using a dynamic IP from your internet provider?

 

[sims]

Do u have your IP addresses fixed on your local netowrk? Are you using a dynamic IP from your internet provider?

1. PIAF is assigned a fixed IP by the router as determined by PIAF's MAC address.

2. From the Internet provider it is a dynamic IP but that has not changed for the last month almost.

 

[wardmundy]

What are we trying to accomplish here? If you've checked your logs and there's no hint there, I think you're at the end of line in finding the source of this problem without some sort of monitoring equipment in place to trap information the next time... if there is a next time. Sorry.

 

[sims]

What are we trying to accomplish here? If you've checked your logs and there's no hint there, I think you're at the end of line in finding the source of this problem without some sort of monitoring equipment in place to trap information the next time... if there is a next time. Sorry.

thanks for asking for this clarification.

i am basically hoping to be able to study logs. just don't know where to look.

should have clarified the question in the first post itself.

 

[rbamburg]

What kind of trunks do you have with this box?

I am testing some boxes myself and getting weird problems.

 

[sims]

two vitelity. one viatalk. and one trunk that a Grandstream HT-503 is connected onto.

i do not know what logs to look at

 

[rbamburg]

One of mine did the same thing. Logs were empty.

I had an Analog Trunk Card in it with 6 fxo ports installed at a client.

I am still testing it to figure out what the problem is but I suspect RAM or Powersupply

 

[Saqqara]

similar crash

I also have PIAF on GPC2, and had a similar crash over the weekend. I don't know where to begin with forensics. The system's not on a UPS at the moment since it's not really in production.

It happened exactly as I submitted a configuration change to my PAP2-NA ATA (changing packet size from default .030 to .010).

Phones never came back up, and couldn't connect by http or putty.

Back in the office, I plugged a monitor in and had no video.

Held the power switch down and it turned off (motherboard listening at least that much). After restarting, asterisk full log contained this:
[Sep 7 04:02:09] VERBOSE[21275] logger.c: Asterisk Event Logger restarted
[Sep 7 04:02:09] VERBOSE[21275] logger.c: Asterisk Queue Logger restarted
[Sep 7 04:02:09] VERBOSE[21275] logger.c: -- Remote UNIX connection disconnected
[Sep 7 05:11:09] DEBUG[3218] sched.c: Request to schedule in the past?!?!
[Sep 7 05:11:09] NOTICE[3218] chan_sip.c: Peer '4436' is now UNREACHABLE! Last qualify: 91
[Sep 7 05:11:19] NOTICE[3218] chan_sip.c: Peer '4436' is now Reachable. (92ms / 2000ms)
[Sep 7 11:03:12] NOTICE[3218] chan_sip.c: Peer '4456' is now Lagged. (2334ms / 2000ms)
[Sep 7 11:03:22] NOTICE[3218] chan_sip.c: Peer '4456' is now Reachable. (212ms / 2000ms)

I'm not sure if the errors here are cause or effect of the crash.

Made a test call to one of the phones on the PAP2-NA (I had left the packet size at .020) and it works. When I get home I'll see if messing with the packet size can reproduce the problem.

I think I need to figure out a way to restart this system remotely. I guess that would have been a good thing about a virtual machine build, had I gone that route (and I might still).

 

[sims]

if you guys figure something out, pl post.

My machine failed one more time

 

[Saqqara]

undetermined

Just to follow up, I can't duplicate the failure by messing with the packet size on the Linksys PAP2-NA and it has run continuously since this one crash (albeit largely unused).

 

[sims]

and for me, it failed yet another time. three times it has happened that i have found that there is no dial tone on the extensions, cannot SSH into the system and the only option is to do a hard physical restart.

 

[RonRussell]

I have deployed two of these systems for clients that wanted a very cheap system and I installed one in my office for testing. Of the three systems I have had 100% failure rate. Everyone of them halts with no odd entries in the logs. I have been working on this issue for about four months. During the same time I have deployed the same distro using Dell PE servers and/or Optiplex desktop systems for four other clients. The Dell installs have a 100% success rate. None of them have locked up at any time. At this time I am replacing all three of the Everex gPC2's. They will be for sale soon at a discount.

 

[sims]

I have had this problem occur five times now. and I still have dropped call problem. not sure if that is because of this hardware or is a software issue.

can you share how much these Optiplex's that you refer to, go for?

It is strange that others are not seeing this problem. Case of bad hardware -- but you have experience with three!

 

[noblenull]

Dell SC440 Servers $199

 

[gregc]

Add me to the list of unknown lockup. Happened yesterday. I was working on it and my SSH session timed out and all the phones went down. When I checked out the monitor on it, it was blank, so I pulled the plug and rebooted. Back up in under two minutes, but annoying.

It has been going for a week without a hitch.

I've added a cron job to do a reboot every tuesday, thursday and saturday early mornings.