SFP-GPON-ONU icon indicating copy to clipboard operation
SFP-GPON-ONU copied to clipboard

DFP-34X-2C2

Open manias21 opened this issue 3 years ago • 21 comments

puedes facilitar la configuracion /var/config/lastgood.xml que utilizaste para movistar?

manias21 avatar Oct 28 '21 00:10 manias21

puedes facilitar la configuracion /var/config/lastgood.xml que utilizaste para movistar?

No, es imposible usarlo para Movistar (con idONT de 0xF20XX...).

zry98 avatar Oct 28 '21 11:10 zry98

firmware para el v2 que te funcione con movistar, probaste alguno?

manias21 avatar Oct 28 '21 11:10 manias21

firmware para el v2 que te funcione con movistar, probaste alguno?

hay un firmware v2? No lo sé, devolví el mío después de descubrir que no es compatible con movistar.

zry98 avatar Oct 28 '21 11:10 zry98

Me a gustado tu publicacion, enorabuena!!!, tienes el firmware a mano y se lo meto?, es por editarlo y pruebo con movistar. gracias.. lo digo porque en mi caso no me deja editarlo, es solo lectura.. si hay algun modo avisa..

manias21 avatar Nov 25 '21 03:11 manias21

Me a gustado tu publicacion, enorabuena!!!, tienes el firmware a mano y se lo meto?, es por editarlo y pruebo con movistar. gracias.. lo digo porque en mi caso no me deja editarlo, es solo lectura.. si hay algun modo avisa..

He agregado algunos firmwares originales que descargué de un grupo chino, incluido uno para el nuevo modelo de ODI, espero que te ayude si arruinaste el tuyo :)

Todavía no lo he investigado, pero la estructura de archivos se ve exactamente igual que los de NATYWISH, TAL VEZ sean intercambiables (es peligroso actualizar un firmware incompatible ya que necesitará abrir el estuche y flash con un programador).

Pueda empaquetar un firmware capaz de editar el idONT más fácilmente (a través de un ejecutable?) cuando tenga tiempo.

zry98 avatar Nov 25 '21 16:11 zry98

hi

I have a DFP-34X-2C2 stick with "STICK3V05" printed on board Looks like it is modern RTL9601D board from ODI

It has two firmwares "V1.7.1-210825" and "V1.0-210702" on board with neither I was able to connect (I have Spanish operator with hex-formatted PLOAM password)

I've tried to kill omci_app(s) processes and then run with -p "$(echo '\x...')" hack but with no luck, it started to switch O2 -> O3 -> O5 (always with "LOID Error")

when it was in O5 with "LOID Error", TX/RX was kind a ok 2.24dbm / -19.78.dbm

  • any hope to make it work?
  • what I am doing wrong?
  • is it ok to kill omci_app and start again? (or I need whole re-build image process to proceed)

p.s. I barely can understand how workaround of -p "<ploam-binary>" can even work, given idea that hex PLOAM password has many zero characters (0x00) inside, and argv in libc runtime is array of pointers to zero-terminated strings (first zero should mean end of parameter string)

vgrebenschikov avatar May 06 '22 09:05 vgrebenschikov

I've tried to kill omci_app(s) processes and then run with -p "$(echo '\x...')" hack but with no luck, it started to switch O2 -> O3 -> O5 (always with "LOID Error")

when it was in O5 with "LOID Error", TX/RX was kind a ok 2.24dbm / -19.78.dbm

Which ISP are you using? It seems you have a wrong LOID and password (not the PLOAM password), Movistar authenticates with only the PLOAM password, but others may also verify MAC address (Orange, Jazztel), LOID & LOID password (vodafone, MasMovil, Yoigo, ...), etc.

  • is it ok to kill omci_app and start again? (or I need whole re-build image process to proceed)

It's OK to kill omci_app processes and run it with the custom parameters in the telnet session, I did that to find this method.

p.s. I barely can understand how workaround of -p "" can even work, given idea that hex PLOAM password has many zero characters (0x00) inside, and argv in libc runtime is array of pointers to zero-terminated strings (first zero should mean end of parameter string)

I think it's working because even though the ocmi_app received the argument as 0xF2021030405678, but the program would eventually pad it with 0x00 to 10 bytes long.

zry98 avatar May 06 '22 12:05 zry98

I've tried to kill omci_app(s) processes and then run with -p "$(echo '\x...')" hack but with no luck, it started to switch O2 -> O3 -> O5 (always with "LOID Error") when it was in O5 with "LOID Error", TX/RX was kind a ok 2.24dbm / -19.78.dbm

Which ISP are you using? It seems you have a wrong LOID and password (not the PLOAM password), Movistar authenticates with only the PLOAM password, but others may also verify MAC address (Orange, Jazztel), LOID & LOID password (vodafone, MasMovil, Yoigo, ...), etc.

Movistar at Barcelona

  • is it ok to kill omci_app and start again? (or I need whole re-build image process to proceed)

It's OK to kill omci_app processes and run it with the custom parameters in the telnet session, I did that to find this method.

ok, fine

p.s. I barely can understand how workaround of -p "" can even work, given idea that hex PLOAM password has many zero characters (0x00) inside, and argv in libc runtime is array of pointers to zero-terminated strings (first zero should mean end of parameter string)

I think it's working because even though the ocmi_app received the argument as 0xF2021030405678, but the program would eventually pad it with 0x00 to 10 bytes long.

Ehh... in mine zeros at the middle 0xF400000057... while in your example, zeroes just like padding at the end :(

vgrebenschikov avatar May 06 '22 20:05 vgrebenschikov

Probably, you know where to get sources of omci_app?

vgrebenschikov avatar May 06 '22 20:05 vgrebenschikov

Probably, you know where to get sources of omci_app?

No, I don't think it's open source. The Realtek SDK is only available to the paid manufacturers.

zry98 avatar May 06 '22 21:05 zry98

Movistar at Barcelona

For Movistar I kept the default loid 123456789 and password 123456.

in mine zeros at the middle 0xF400000057...

that's very strange! all the hex idONT (ploam password) I have seen are in the format of 0xFYYYYMMDDXXXXXXXXXXX, where YYYYMMDD I think is the date the fiber cable got installed.

zry98 avatar May 06 '22 21:05 zry98

Los nuevos ploam ya son F4xxxxxxx. y si, hay un firmware que chuta ya con movistar, https://github.com/Anime4000/RTL960x/blob/main/Firmware/DFP-34X-2C2/M114_sfp_ODI_220304.tar

Actualiza con este firmware por web, despues por terminal telnet aplica flasherase_all /dev/mtd3, reboot, entra de nuevo por web y aplica el ploam, ya podras conseguir el 05, despues para conocer la vlan que vincular a tu pppoe es el comando 84 que te listara las vlan disponibles y una de ellas es la correcta, la vlan6 te aseguro que no sera la tuya proque los mibs no estan ajustados para ello.

manias21 avatar May 07 '22 09:05 manias21

omcicli mib get 84

manias21 avatar May 07 '22 09:05 manias21

Los nuevos ploam ya son F4xxxxxxx. y si, hay un firmware que chuta ya con movistar, https://github.com/Anime4000/RTL960x/blob/main/Firmware/DFP-34X-2C2/M114_sfp_ODI_220304.tar

Actualiza con este firmware por web, despues por terminal telnet aplica flasherase_all /dev/mtd3, reboot, entra de nuevo por web y aplica el ploam, ya podras conseguir el 05, despues para conocer la vlan que vincular a tu pppoe es el comando 84 que te listara las vlan disponibles y una de ellas es la correcta, la vlan6 te aseguro que no sera la tuya proque los mibs no estan ajustados para ello.

with versions 220414 or 220304 I can put hex password ... and even get O5 ...

but PPPoE does not established (no response to ours PPPoE requests)

Same time other VLANs are works somehow

  • DHCP works on VoIP VLAN (3), but SIP calls are not work
  • UDP Multicasts on IPTV VLAN (2) ~ 30 Mbps of incoming traffic
  • but on both interfaces no RIP routes

according to what I see in tcpdump - frames get from SFP without VLAN tags !!

Movistar ... no idea what is wrong (same setup works fine with UFiber LOCO)

Looks like SFP somehow eat VLAN tags, probably to one side

Let me try omcicli mib get 84

vgrebenschikov avatar May 07 '22 20:05 vgrebenschikov

That is strange:

# omcicli mib get 84
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
VlanTagFilterData
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
=================================
EntityID: 0x03
FilterTbl[0]: PRI 0,CFI 0, VID 290
FwdOp:  0x10
NumOfEntries: 1
=================================
=================================
EntityID: 0x04
FilterTbl[0]: PRI 0,CFI 0, VID 6
FwdOp:  0x10
NumOfEntries: 1
=================================
=================================
EntityID: 0x05
FilterTbl[0]: PRI 0,CFI 0, VID 3
FwdOp:  0x10
NumOfEntries: 1
=================================
#

for UFiber LOCO, following schema of VLANs works fine:

Columns: NAME, MTU, ARP, VLAN-ID, INTERFACE
#   NAME        MTU  ARP      VLAN-ID  INTERFACE
0 R gpon.inet  1500  enabled        6  gpon
1 R gpon.iptv  1500  enabled        2  gpon
2 R gpon.voip  1500  enabled        3  gpon

vgrebenschikov avatar May 07 '22 21:05 vgrebenschikov

Tramita la prueba de montar el pppoe con la vlan 290 o la vlan 6.

manias21 avatar May 07 '22 21:05 manias21

Tramita la prueba de montar el pppoe con la vlan 290 o la vlan 6.

Yes, PPPoE on VLAN 290 started to work, but completely unclear why schema was changed if I compare with what I see on UFiber LOCO ...

(no VLAN 290 - internet, VLAN 6 - VoIP)

vgrebenschikov avatar May 07 '22 21:05 vgrebenschikov

aaaaaaaaaa... si, es muy normal, los mibs estan generados de ese modo en los firmwares.

manias21 avatar May 07 '22 21:05 manias21

Hm... I've thought that VLAN schema is defined on provider side, not by firmware

vgrebenschikov avatar May 07 '22 22:05 vgrebenschikov

si, el isp lo suministra, pero el firmware lo adapta a su configuracion propia de mibs.

manias21 avatar May 07 '22 22:05 manias21

si, el isp lo suministra, pero el firmware lo adapta a su configuracion propia de mibs.

Have you succeeded to get VoIP and IPTV with the stick?

vgrebenschikov avatar May 08 '22 12:05 vgrebenschikov