RouterTech Firmware 2.93 Discussion

[thechief]

I would imagine that pppoa is less troublesome. Two hours is probably to great an interval for a dodgy line. Try 15 minutes instead!

[mysticalos]

I found a way to reproduce this bug. PPP fails if you go to setup, then go to your ppp connection, in my test cases, pppoa. Then hit disconnect button, followed by hitting connect button quickly. The router does disconnect fine, even reconnects fine, but despite having reconnected, the ppp daemon fires up and tries to continue connecting when you're already connected, then spams errors that process is busy (cause you're already connected). The problem, is that even though you were already connected, because router doesn't realize this as soon as automation kicks in you lose your connection anyways and now you got a hung connection that's still connected, while router indefinitely tries to establish a connection over the hung connection. An annoying bug to say the least, but it has to be fairly recent bug cause in earlier versions i always used disconnect/connect buttons manually in setup if I lost a connection and didn't wanna reboot the router. Whether this can be fixed I dunno, it'd probably need fixing in busybox no?

EDIT: I discovered i could avoid the above by hitting disconnect, then waiting about 30 seconds before hitting connect again and it doesn't get messed up.

Checksync.sh no longer shows in log? I noticed when i enabled it to run every 10 minutes i don't see anything in log showing it did it's job. i had to wait for connection to drop before i knew it was working.

[thechief]

An annoying bug to say the least, but it has to be fairly recent bug

We have been on this version of pppd since September 2009.

cause in earlier versions i always used disconnect/connect buttons manually in setup if I lost a connection and didn't wanna reboot the router. Whether this can be fixed I dunno, it'd probably need fixing in busybox no?

Who knows? The problem could be in pppd, dnsmasq, busybox or in the wait periods in the config.

EDIT: I discovered i could avoid the above by hitting disconnect, then waiting about 30 seconds before hitting connect again and it doesn't get messed up.

That is what to do. Disconnecting and connecting immediately without giving the system some time is a recipe for trouble.

Checksync.sh no longer shows in log?

No. It wasn't supposed to be showing before. One of those quiet busybox fixes.

[mysticalos]

Is there any way to add dslreports dyn dns support?

http://www.dslreports.com/faq/monitorfaq

With their line monitoring features setting up a dyn dns helps you keep line monitor up even through ip changes. routertech already supports dyn dns but it just doesn't have the dslreports preset coded in. I'm wondering if this wouldn't be too difficult to add? I think it'd be a useful feature.

Another question, my log is full of

May 25 19:34:37 | bad password attempt for 'root' from 189.68.174.203:4657

Doing a traceroute on IP revealed it was a telephone/ad/sale service IP. I had to turn the firewall off for line monitoring. I'll probably just have to turn firewall back on and figure out what IP range is used by line monitor and add ip filters for em.

[mstombs]

New modules can be added to dyndns, but can't see anybody bothering for such a commercial service!

Don't allow remote access if you don't understand what you are doing! You should be able to allow password protected services to selected external IP ranges - I wonder if this why you have high latency?

[mysticalos]

i don't allow remote access, unless it's on by default? and the firewall was on until yesterday.

edit: just checked, remote access is turned off.

[mstombs]

I assume remote access restrictions require the firewall to be enabled, so web interface, telnet and ssh would be reliant on simple passwords only! Not too sure what an attacker could do, but the enhanced feature set of RouterTech routers is a disadvantage of RouterTech firmware. Most likely attack would be to use the router to send out span through your ISP account?

[mysticalos]

I did a save all, after changing modulation, exactly like i did before, and like before, the router got bricked again, just like it did under adam2, only this time under psp boot. Are you sure there isn't a big nasty bug in there somewhere? I'm on ISP bad again to download PC tool, which at least i will be able to use this time since there is a vista one for psp bootloader. :\ I do change the modulation 2-3 times a day as G.Dmt gets the best margins and speeds, ADSL2+ gets more stable connection (don't ask me why, since just using that over G.dmt makes me lose 3 off up and down margin) and ADSL2 regular is a balance in between (and actually is waht's chosen when it's set to multi mode). Is manually changing modulation all the time though a bad thing? saving a new config multiple times a day lead to eventual bricking/corruption? I'll see what environment looks like in a bit here.

Edit:

OK, using PC tool i determined env was fine, so was firmware, only the config space was corrupt, restored router by uploading a new config. It seemed for whatever reason, as soon as autoload timeout ran, the router was losing control of all the ethernet ports. Very strange. The LEDs would show traffic but it was almost as if they were firewalled and on computer end it would just think there was nothing there. Wonder how changing modulation would randomly kill the ethernet ports, but i do know Dlink 604Ts in past had quirky ethernet switch behaviors so maybe that's it? Anyway back up and running again

[mstombs]

I'm afraid Ethernet ports more likely to be hardware issue - have you checked the PSU. There is a strange problem with some 2640Ts QOS reconfigures the switch?

There is a potential vulnerability with any flash writes - the power must not fail! Reading and writing the env+config could be an issue, but very unusual for just the config to be corrupt. The modulation is stored in the env, the env+config occupy aingle 64kB flash sector and any change means the whole sector is erased and everything re-written.

Regular writes that you describe are not typical - I suggest you should try not to be too hasty in multiple reads/ writes¬

[eskasi]

I have read about to use 30 secs as keepalive period.
Sorry for my english

Hi all!
I have given life back to a router from a jobmate, an old d-link 504T.

I have succesfully flashed to routertech firmware v2.93 (4mb flash, adam2, 4 ports switch)....but i feel this router is unresponsive.

I have setup my dsl connection, dns, enable remote access and i am here, at job, doing a ssh to home with Putty.
The connection breaks very often: dsl carrier is down; "ds margin"=9 and "us margin"=8 (ds attenuation=0, us attenuation=8). Putty is configured to send keep alive packets every 300 secs.

I wanted to use my upload capabilities, so i had enabled the annex M....but i feel a bit deceived...because the connection is very unstable.

I also have noticed that when i use the router web interface, the other activity of the router...collapses. In fact, i wanted to disable the http daemon from wan side, but the enable mark does not work for me.

I only wanted to give a second life to this old router...should i continue with my efforts or directly throw it into trash??

Regards.

[mstombs]

Yes the ssh server daemon requires a keepalive of about 30 seconds - this is so that broken connections do get cleaned up.

Should have nothing to do with your other problems, which seem many and varied, please provide more info - and this should probably be split into a separate thread if not specifically about 2.93

Obtaining help with RouterTech firmwares

[eskasi]

Yes the ssh server daemon requires a keepalive of about 30 seconds - this is so that broken connections do get cleaned up.

Should have nothing to do with your other problems, which seem many and varied, please provide more info - and this should probably be split into a separate thread if not specifically about 2.93

Obtaining help with RouterTech firmwares

Thank you!
I will post it in a separate thread when i get some more information about anyone with this router. In fact, i have tried an official firmware and the web interface is also very sluggish! Maybe the fault is not the firmware...

[felixmarra]

I enabled "Enable Cron (cron_enable)" and "Enable Memory optimisation (memchk.sh - needs "Cron")" with 1 20 parameters (memory optimization every 20 minutes, right?) but it doesn't work according to the System Log.

With the 2.92 it worked (but I left only "1" in parameters and memory optimization was every hour). Any idea?


Apr 8 09:23:11 | Error getting mac from BootLoader enviroment for wlan0
Apr 8 09:23:11 | Using default mac address: 00.50.F1.12.12.10
Apr 8 09:23:11 | Use BootLoader command:
Apr 8 09:23:11 | setenv mac_ap xx.xx.xx.xx.xx.xx
Apr 8 09:23:11 | to set mac address
Apr 8 09:23:11 | Valid Configuration Tree
Apr 8 09:23:11 | NTP Polling Timer for DHCP Started succesfully.
Apr 8 09:23:11 | DSL Polling Timer Started succesfully.
Apr 8 09:23:11 | PSP Boot environment Modem Modulation Change: 0x1
Apr 8 09:23:11 | Firewall NAT service started
Apr 8 09:23:12 | Bridge Created: br0
Apr 8 09:23:14 | Bridge VLAN0 add eth0
Apr 8 09:23:14 | Bridge VLAN AUTO OFF.
Apr 8 09:23:14 | Bridge VLAN0 add eth1
Apr 8 09:23:14 | Bridge VLAN AUTO OFF.
Apr 8 09:23:14 | Bridge VLAN0 add eth2
Apr 8 09:23:14 | Bridge VLAN AUTO OFF.
Apr 8 09:23:14 | Bridge VLAN0 add eth3
Apr 8 09:23:14 | Bridge VLAN AUTO OFF.
Apr 8 09:23:15 | Bridge Created: br1
Apr 8 09:23:16 | Bridge Created: br2
Apr 8 09:23:19 | Bridge Interface Added: eth0
Apr 8 09:23:21 | Bridge Interface Added: eth1
Apr 8 09:23:23 | Bridge Interface Added: eth2
Apr 8 09:23:25 | AP Driver configuration successful
Apr 8 09:23:29 | AP IS UP
Apr 8 09:23:30 | Bridge Interface Added: eth3
Apr 8 09:23:31 | USB is disabled
Apr 8 09:23:31 | DSL Carrier is down
Apr 8 09:23:32 | Bridge Interface Added: wlan0
Apr 8 09:24:01 | DSL Carrier is up
Apr 8 09:24:01 | sar read trained mode (5)(ADSL_2plus)
Apr 8 09:24:01 | ---}}} Start of connection delayed for 2 sec
Apr 8 09:24:01 | PPPoA launch delay.
Apr 8 09:24:03 | PPPoA Launch after conn delay timeout ...
Apr 8 09:24:04 | pppd 2.4.4 started by root, uid 0
Apr 8 09:24:04 | Connect: ppp0 {--}
Apr 8 09:24:09 | PPPD Successfully Started
Apr 8 09:24:10 | CHAP authentication succeeded
Apr 8 09:24:10 | local IP address 62.xxx.xxx.xxx
Apr 8 09:24:10 | remote IP address 62.xxx.xxx.xxx
Apr 8 09:24:10 | PPPoA Connect with IP Address 62.xxx.xxx.xxx
Apr 8 09:24:10 | PPPoA Connection Successfully Established
Apr 8 09:24:10 | PPPoA Connect with Gateway IP Address: 62.xxx.xxx.xxx
Apr 8 10:24:13 | DDNS noip: Update Successful ip 62.xxx.xxx.xxx from ppp0
Jun 9 12:08:28 | onconnectWAN: Crond is already running.
Jun 9 12:09:05 | time disparity of 89385 minutes detected



EDIT: changed the parameters to "1" as it was before but nothing changed, still not ram optimization...

[thechief]

I enabled "Enable Cron (cron_enable)" and "Enable Memory optimisation (memchk.sh - needs "Cron")" with 1 20 parameters (memory optimization every 20 minutes, right?) but it doesn't work according to the System Log.

What makes you think it doesn't work? The log entries have stopped (one of those quiet busybox fixes), but they should not have been there in the first place!

[felixmarra]

What makes you think it doesn't work? The log entries have stopped (one of those quiet busybox fixes)

DOH!

but they should not have been there in the first place!

how to see if it is working now?

I didn't know it was a bug and i was used with it.