Gargoyle Forum

First of all let me say that the firmware is great and very stable.
I'm running version 1.0.10 in a Fonera+ with QOS and DDNS activated.
I've also been able to make a heartbeat at every 29 min in the fon network.

But, i've noticed that about every 12h the router reboots itself - I notice this when the wireless and cable connection disappears and get back after around 1 min and the router reports a uptime of 1 minute !
I've never seen an uptime bigger than 12h.

It is not a major issue for normal browsing, but if gaming in the playstation network when the reboot happens it is a little bit annoying...

I've not found any job in the cron tabs that could produce this issue.

Is this normal ? Is it really 12 hours ?

Thank you all

I also have a Fonera+ and have been testing Gargoyle on it
for quite a while now. I've never seen it auto reboot at the
12 hour mark. I've had uptimes of 1-2 weeks for some builds.

The Fonera+ can easily run out of RAM though, which might
cause it to crash and reboot.

1) Are you running torrents?
2) Have you limited the # of connections?
3) Have you changed the timeouts on connections?
4) Do you need all the monitoring?

I've done extensive testing with the Fonera+ with torrents.
I've found that with a few hundred incoming connections,
EVEN IF you are NOT maxing out the bandwidth either up/down
will definitely bring the router to a crawl.
Just browsing web pages while the above is happening will
take upto 1 minute to load some web pages.

Torrents are particularly horrible on the Fonera+

Hi:

Yes. I'm using torrents and using QOS to limit p2p as slow.
What is your sugggestion about number of connections ?

Should I also limit connections in the bittorent client (transmission running in a fonera 2) ?

I'm using

1024 max connections.
TCP timeout = 600 secs
UDP timeout = 180 secs

Ok...

Here's what I did for testing:

I have a max download of 10mbps and max upload of 768kbps

I download 6 Linux dvd images via torrent

I had all 6 running at once, and I limited each of them
to 64kbps upload, no limit on connections.

So, 6 x 64kbps = 384kbps max upload

When all 6 were going maxed, they had around 200-ish
connections going at once for all 6 torrents.

So, I still had PLENTY of bandwidth, but the Fonera 2201+
could just not keep up. Any internet browsing was just
slow as hell and couldn't do anything else.

You can try and play around with settings, but I'd
recommend the following settings and then adjust:

no more than 1/2 of your upload bandwidth
keep it under 200 connections
no more than 5 torrents at once
TCP timeout of 300 or 360 seconds

Try that out and then see how your other internet use works.

Hi:

Thanks for your suggestions.
I have 10Mbit download and 1Mbit upload.

I've just change in the router:
Max Connections: 1024
TCP Timeout: 120 seconds
UDP Timeout: 120 seconds
Defined in QOS downolad , the port 51413 as "slow"
maximum download at 9000 Kbps

In my torrent application, I've limited download to 200kbit and upload to 20 Kbit.
I usually do not have more than 2 torrents working.

It's been running well for the past hours.
Will post the results in a couple of days.

Kind Regards

The system has been working fine.
Today , no torrent downloading.

I just went to the gargoyle interface, and right now, (15h33m, WET)
At the status window:
Uptime:0 days, 3 hours, 33 minutes

The router rebooted at 12h00 sharp !
Just like yesterday !!

That is really really weird.

Since you're working off a Fonera+ like I am, I would recommend
trying to completely re-flash the device and wipe all settings.

Goto the System tab and backup your config.

Reboot the router and Telnet into the Redboot console.

Do the command: fis init -f to completely wipe the flash memory.

Then either manually load or use Fon-Flash to reload 1.0.10
Restore the config and see if any reboot occurs.

The Gargoyle interface has no way to input cron jobs, so it would
have been something you manually entered in to involve a reboot.

Here's my uptime so far:

Well...
It's the same:
It randomly restarts but only at "full-hours" (i believe it is the correct english term), like 12:00, 4:00, 18:00, etc...

From my understanding the only thing that runs at this time is the ntp demon. right ?

I've just restarted the router. If it happens again, I will "kill" ntpclient and wait ...

Try the new version 1.0.11. The bandwidth monitor backs up data every 4 hours, which may help explain what was going on for you. Cron was scheduling a whole bunch of commands to be executed (to do the backup for each IP) at those times -- it's possible if you had a lot of IPs defined to be monitored, or you were really low on RAM, all these concurrent jobs were eating up the RAM and the router was rebooting. The new version executes a script from Cron which runs the necessary commands sequentially instead of in parallel so they don't eat up too much memory.