Last time i use this OpenWrt Attitude Adjustment 12.09-rc1 - OpenWrtID / LuCI 0.11 Branch (0.11+svn9599) all fine but after upgrade to Back window or Huntsman cannot access to Internet.

I follow this step to setting RC1
- login 192.168.1.1 ( you will see LuCI Gui interface )
- click login
- go to System
- go to Administration
- key in your new password ( For web Gui )
- re key in your same password again
- click save and apply
- go to Network
- click Edit on Wan ( 3G wan )
- select UMTS/GPRS/ED-VO protocol
- click switch protocol
- Modem device ( ZTE modem = /dev/ttyUSB1 , Huawei modem = /dev/ttyUSB0 <- last is zero )
- service type : select UMTS only or UMTS/GPRS
- Key in your ISP APN , ( Unet, Maxisbb, bbnet, celcom3g, u3g, my3g and so on )
- Go to ADvance setting
- Modem init time out: 5 ( key in )
- Use gateway Metric : 0
- Tick the box of : if uncheck, No default route is configure
- Use custom DNS server : 8.8.8.8 click the icon at the end
- second roll: 8.8.4.4
- LCP echo failed Threshold : 0
- LCP echo interval : 5
- Inactivity timeout: 0
- select Wifi
- click Edit
- select your wifi prefer channel 1~11
- enter your wifi name in ESSID
- go to Wireless security
- encryption : WPA2-PSK
- key : wifi password
- click save and apply
- then click enable to on your wifi ( if u need it )
- now go to System
- Reboot
- click Performs reboot
- plug in your Usb modem after
- wait for reboot finish and enjoy

Why?

I use V2 3420 + ZTE MF680 modem

TQ

after mine hang, the usb modem is not recognizable, and no matter how many time reboot still same, 

wanna change back to stock firmware..... ...

change to E372...be like this...

Thanks for the sharing...i do facing the same issue if i am using "Enabled with Restart" for Connection Keep Alive
May i know is it you use this command:

* * * * * /bin/ping -q -s 1 -c 60 8.8.8.8 & > /dev/null

as your own scheduled tasks to monitor the connection and restart if necessary?

i see...well, i will try to use the command and settings that you mention above and see what happen next.

To be honest, with the Huntsman's feature, i feel reluctant to use the AA version 

that command is to keep the connection alive by pinging Google DNS every minute but doesn't do anything else.

I'm using this:
*/5 * * * * /bin/ping -c 10 openwrt.org || reboot

It will ping openwrt.org every 5 minutes with 10 tries and will run the reboot command if one or more out of the 10 pings failed. You can change the schedule to a shorter/longer time or a different address to ping depending on your preference.

I can't guarantee that it will fix the issue though 'cause we'll need more time to test it out. However, at least the logic of it make sense. 

latest update on my tests:
It seems that adding the scheduled task doesn't solve the issue. From my observation, when the issue happen, none of the scheduled tasks run. I'm assuming that cron service somehow froze up.

Here's what I'm testing now:
Take out the original scheduled task added in ROOTer firmware to keep the connection alive but leave my own tasks intact. I restart the cron service on the safe side to ensure that it will run the new config.
*/2 * * * * /etc/root  <--- deleted this

It seems that BW 2013.04.04 is much stable after all ?

The later Black Widow and Huntsman builds have introduced instability for some modems.

We are currently working back through the ROOter additions to isolate the problem(s).

Your testing has been invaluable as it identified the modem instability and the keep alive function failure which had not beenreported elsewhere. 

... Back with an update when there is more news.

Cheers

saw this when connect n using the previous password with putty

should me my hard reset dint work n the password still exist right?

this is the msg i get from putty

"root@akping:~# /etc/init.d/uhttpd start
Failed to load ustream-ssl library: File not found "

am still using the above AA version for my mr3040V1 ,. so far is the most stable and best in consistency openwrt,..



reason to support my statement above is my load normally always between 0.21~0.25 only,..

which when i use blackwidow or huntsman, my loads is between 0.79~0.98 
router load when running torrent 0.21~0.25
[attachmentid=3407212]
router load when not running torrent 0.3~0.10
[attachmentid=3407218]
and when the router load running too high, sometimes the webinterface 192.168.1.1 become unresponsive and the wireless connection get affected like request time out too,.. 

so before anyone of you think is the temperature issue for the wireless connection, this is how i setup my MR3040 V1 + Huawei E1820 ( with usb mini usb table fan blowing )
so it is not the temperature issue

[attachmentid=3407220]

Care to share with me your cron settings? As i still having the same issue 

so a "new virgin image" ?

to be flashed on which version of the Huntsman ?

thanks

 

i did test with the hutnsman 4-19, my averages load was always very high even there is no internet traffic loading,. 

Both my 1043ND and MR3420 are running HM 04-14, only MR3420 fakap on me, whereas 1043ND works flawlessly.

Maybe 1043ND have a bigger flash (8MB) , hence less problem detected.

 

My 1043ND session starts at 11.30pm > 10.30am , not even an IP change sometimes............mostly streaming sky.fm  classical jazz 

No idea what happened, i try to downgrade from 04-23 to 04-19 then upgrade again to 04-30, surprisingly now its working, no more disconnect

BTW sierra modems only can use DCHP client? i tried UMTS/GPRS/EVDO and PPP mode but end up get no connection.

just tested the hunstman 4-30 with my MR3040V1

it can auto detect and dial even though the network status showing wrong port or detail display is haywire,..

[attachmentid=3417082]
[attachmentid=3417083]