open5gs/docs/_pages/faq.md at 80199a3c877557cbb226b249baa57a45be341897

mirror of https://github.com/open5gs/open5gs.git synced 2026-05-05 23:37:22 +00:00

Sukchan Lee f7ef0f3aab FAQ update

2019-06-02 00:50:46 +09:00

16 KiB

Raw Blame History

layout	title	permalink
page	FAQ	/faq/

Is it possible to setup IP/NAT table along with Docker?

The following is the default docker IP/NAT table.

$ sudo iptables-save
# Generated by iptables-save v1.6.1 on Sat Jun  1 23:43:50 2019
*nat
:PREROUTING ACCEPT [74:4820]
:INPUT ACCEPT [4:208]
:OUTPUT ACCEPT [49:3659]
:POSTROUTING ACCEPT [49:3659]
:DOCKER - [0:0]
-A PREROUTING -m addrtype --dst-type LOCAL -j DOCKER
-A OUTPUT ! -d 127.0.0.0/8 -m addrtype --dst-type LOCAL -j DOCKER
-A POSTROUTING -s 172.17.0.0/16 ! -o docker0 -j MASQUERADE
-A DOCKER -i docker0 -j RETURN
COMMIT
# Completed on Sat Jun  1 23:43:50 2019
# Generated by iptables-save v1.6.1 on Sat Jun  1 23:43:50 2019
*filter
:INPUT ACCEPT [651:514108]
:FORWARD DROP [70:4612]
:OUTPUT ACCEPT [590:63846]
:DOCKER - [0:0]
:DOCKER-ISOLATION-STAGE-1 - [0:0]
:DOCKER-ISOLATION-STAGE-2 - [0:0]
:DOCKER-USER - [0:0]
-A FORWARD -j DOCKER-USER
-A FORWARD -j DOCKER-ISOLATION-STAGE-1
-A FORWARD -o docker0 -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT
-A FORWARD -o docker0 -j DOCKER
-A FORWARD -i docker0 ! -o docker0 -j ACCEPT
-A FORWARD -i docker0 -o docker0 -j ACCEPT
-A DOCKER-ISOLATION-STAGE-1 -i docker0 ! -o docker0 -j DOCKER-ISOLATION-STAGE-2
-A DOCKER-ISOLATION-STAGE-1 -j RETURN
-A DOCKER-ISOLATION-STAGE-2 -o docker0 -j DROP
-A DOCKER-ISOLATION-STAGE-2 -j RETURN
-A DOCKER-USER -j RETURN
COMMIT
# Completed on Sat Jun  1 23:43:50 2019

Create newtables file as below.

$ diff -u oldtables newtables
--- oldtables	2019-06-01 23:43:50.354974226 +0900
+++ newtables	2019-06-01 23:44:16.110931684 +0900
@@ -8,6 +8,7 @@
 -A PREROUTING -m addrtype --dst-type LOCAL -j DOCKER
 -A OUTPUT ! -d 127.0.0.0/8 -m addrtype --dst-type LOCAL -j DOCKER
 -A POSTROUTING -s 172.17.0.0/16 ! -o docker0 -j MASQUERADE
+-A POSTROUTING -s 45.45.0.0/16 ! -o pgwtun -j MASQUERADE
 -A DOCKER -i docker0 -j RETURN
 COMMIT
 # Completed on Sat Jun  1 23:43:50 2019
@@ -26,6 +27,10 @@
 -A FORWARD -o docker0 -j DOCKER
 -A FORWARD -i docker0 ! -o docker0 -j ACCEPT
 -A FORWARD -i docker0 -o docker0 -j ACCEPT
+-A FORWARD -o pgwtun -m conntrack --ctstate RELATED,ESTABLISHED -j ACCEPT
+-A FORWARD -o pgwtun -j DOCKER
+-A FORWARD -i pgwtun ! -o pgwtun -j ACCEPT
+-A FORWARD -i pgwtun -o pgwtun -j ACCEPT
 -A DOCKER-ISOLATION-STAGE-1 -i docker0 ! -o docker0 -j DOCKER-ISOLATION-STAGE-2
 -A DOCKER-ISOLATION-STAGE-1 -j RETURN
 -A DOCKER-ISOLATION-STAGE-2 -o docker0 -j DROP

And then, apply newtables as below.

$ sudo iptables-restore < newtables

The above operation is the same as described in the following manuals.

### Check IP Tables
$ sudo iptables -L
Chain INPUT (policy ACCEPT)
target     prot opt source               destination

Chain FORWARD (policy ACCEPT)
target     prot opt source               destination

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination

### Check NAT Tables
$ sudo iptables -L -t nat
Chain PREROUTING (policy ACCEPT)
target     prot opt source               destination

Chain INPUT (policy ACCEPT)
target     prot opt source               destination

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination

Chain POSTROUTING (policy ACCEPT)
target     prot opt source               destination

### Enable IPv4 Forwarding
$ sudo sh -c "echo 1 > /proc/sys/net/ipv4/ip_forward"

### Add NAT Rule
$ sudo iptables -t nat -A POSTROUTING -s 45.45.0.0/16 ! -o pgwtun -j MASQUERADE

How to use a different APN for each PGW

By default, MME selects the PGW as the first PGW node. To use a different APN for each PGW, specify gtpc.apn as the APN name. If the HSS uses WebUI to set the PGW IP for each UE, you can use a specific PGW node for each UE.

See the following example.

### For reference, see `nextepc.conf`
#------------------------ MME --------------------------
#
#  o Two PGW are defined. 127.0.0.3:2123 is used.
#    [fe80::3%lo]:2123 is ignored.
#    gtpc:
#      - addr: 127.0.0.3
#      - addr: fe80::3%lo
#
#  o One PGW is defined. if prefer_ipv4 is not true,
#    [fe80::3%lo] is selected.
#    gtpc:
#      - addr:
#        - 127.0.0.3
#        - fe80::3%lo
#
#  o Two PGW are defined with a different APN.
#    - Note that if PGW IP for UE is configured in HSS,
#      the following configurion for this UE is ignored.
#    gtpc:
#      - addr: 127.0.0.3
#        apn: internet
#      - addr: 127.0.0.5
#        apn: volte
#
#  o If APN is omitted, the default APN uses the first PGW node.
#    gtpc:
#      - addr: 127.0.0.3
#      - addr: 127.0.0.5
#        apn: volte
#

The IP address of the UE can also use a different UE pool depending on the APN.

### For reference, see `nextepc.conf`
#
#  <UE Pool>
#
#  o IPv4 Pool
#    $ sudo ip addr add 45.45.0.1/16 dev pgwtun
#
#    ue_pool:
#      addr: 45.45.0.1/16
#
#  o IPv4/IPv6 Pool
#    $ sudo ip addr add 45.45.0.1/16 dev pgwtun
#    $ sudo ip addr add cafe:1::1/64 dev pgwtun
#
#    ue_pool:
#      - addr: 45.45.0.1/16
#      - addr: cafe:1::1/64
#
#
#  o Specific APN(e.g 'volte') uses 45.46.0.1/16, cafe:2::1/64
#    All other APNs use 45.45.0.1/16, cafe:1::1/64
#    $ sudo ip addr add 45.45.0.1/16 dev pgwtun
#    $ sudo ip addr add 45.46.0.1/16 dev pgwtun
#    $ sudo ip addr add cafe:1::1/64 dev pgwtun
#    $ sudo ip addr add cafe:2::1/64 dev pgwtun
#
#    ue_pool:
#      - addr: 45.45.0.1/16
#      - addr: cafe:1::1/64
#      - addr: 45.46.0.1/16
#        apn: volte
#      - addr: cafe:2::1/64
#        apn: volte
#
#  o Multiple Devices (default: pgwtun)
#    $ sudo ip addr add 45.45.0.1/16 dev pgwtun
#    $ sudo ip addr add cafe:1::1/64 dev pgwtun2
#    $ sudo ip addr add 45.46.0.1/16 dev pgwtun3
#    $ sudo ip addr add cafe:2::1/64 dev pgwtun3
#
#    ue_pool:
#      - addr: 45.45.0.1/16
#      - addr: cafe:1::1/64
#        dev: pgwtun2
#      - addr: 45.46.0.1/16
#        apn: volte
#        dev: pgwtun3
#      - addr: cafe:2::1/64
#        apn: volte
#        dev: pgwtun3
#

How to connect MongoDB server

If you do not start MongoDB, you will get the following error:

➜  nextepc git:(master) ✗ ./test/testsimple  
04/09 15:42:34.817: [core] ERROR: Failed to conect to server [mongodb://localhost/nextepc] (context.c:326)  
04/09 15:42:34.817: [core] ERROR: app_initialize() failed (basic/abts-main.c:91)  
s1ap_message_test   : -04/09 15:42:34.830: [s1ap] ERROR: Failed to decode S1AP-PDU[-1] (s1ap_encoder.c:41)  
04/09 15:42:34.830: [core] FATAL: ogs_log_vprintf: Assertion `domain' failed. (ogs-log.c:347)

You can start MongoDB using systemctl.

$ sudo systemctl start mongodb

Failing to run `./nextepc-epcd`

You might be getting the following error after running ./nextepc-epcd.

➜  nextepc git:(master) ./nextepc-epcd  
04/09 15:41:02.600: [core] INFO: NextEPC daemon start (main.c:169)  
04/09 15:41:02.601: [core] ERROR: CHECK PERMISSION of Installation Directory... (application.c:144)  
04/09 15:41:02.601: [core] ERROR: Cannot create PID file:`/home/acetcom/Documents/git/open5gs/nextepc/install/var/run/nextepc-epcd/pid` (application.c:145)  
04/09 15:41:02.601: [core] WARNING: log_pid: should not be reached. (application.c:146)  
/home/acetcom/Documents/git/open5gs/nextepc/lib/ogslib/src/core/.libs/libogscore-1.0.so.0(ogs_abort+0x2b)[0x7f9d5d26d71b]  
/home/acetcom/Documents/git/open5gs/nextepc/.libs/nextepc-epcd(+0x9606)[0x563a4ba23606]  
/home/acetcom/Documents/git/open5gs/nextepc/.libs/nextepc-epcd(+0x8640)[0x563a4ba22640]  
/home/acetcom/Documents/git/open5gs/nextepc/.libs/nextepc-epcd(+0x81f3)[0x563a4ba221f3]  
[1]    9635 abort (core dumped)  ./nextepc-epcd

You should perform the installation process.

$ make install

I have some error when running `./test/testcomplex`

Did you see the following error after executing testcomplex?

➜  nextepc git:(master) ✗ ./test/testcomplex  
s1setup_test        : SUCCESS  
attach_test         : -Line 134: Condition is false, but expected true  
\04/09 15:49:09.285: [esm] FATAL: esm_handle_pdn_connectivity_request: Assertion `SECURITY_CONTEXT_IS_VALID(mme_ue)' failed. (esm_handler.c:29)  
/home/acetcom/Documents/git/open5gs/nextepc/lib/ogslib/src/core/.libs/libogscore-1.0.so.0(ogs_abort+0x2b)[0x7f608518271b]  
/home/acetcom/Documents/git/open5gs/nextepc/test/.libs/testcomplex(+0x92121)[0x55dc9e274121]  
/home/acetcom/Documents/git/open5gs/nextepc/test/.libs/testcomplex(+0x4f5b9)[0x55dc9e2315b9]

Remove all subscriber information using MongoDB Client

$ mongo
> db.subscribers.find()  ### Check the test subscriber
> db.subscribers.drop()  ### Remove all subscriber
> db.subscribers.find()  ### Check that all subscribers are empty

Kill all processes.

$ ps -ef | grep testcomplex
$ ps -ef | grep nextepc
$ sudo pkill -9 testcomplex
$ sudo pkill -9 nextepc-epcd ...

Execute testcomplex

$ ./test/testcomplex

My eNB does not support IPv6.

Your eNodeB don't have to support IPv6.

If the sgw.gtpu configuration does not have an IPv6 address, the eNodeB can use IPv4 to connect to the MME and SGW. If the sgw.gtpu setting has an IPv6 address, you can disable the IPv6 address as shown below.

parameter:
    no_ipv6: true

Note: This parameter no_ipv6 is only applied to EPC Elements such as MME, SGW, and so on. The parameter no_ipv6 does not affect to UE. So, IPv6-enabled UE can connect to NextEPC LTE network. {: .notice--warning}

Unable to add new user by WebUI

It might be a CSRF token mismatch error. Please, delete corresponding cookies, cache, session data etc.

Change Domain from localdomain to anything else

You should configure the domain name on your computer. Otherwise, freeDiameter raise an error.

How many of UEs can NextEPC support?

See the [lib/base/types.h](https://github.com/{{ site.github_username }}/nextepc/blob/master/lib/base/types.h).

#define MAX_NUM_OF_ENB              128
#define MAX_NUM_OF_UE               128
#define MAX_NUM_OF_SESS             4
#define MAX_NUM_OF_BEARER           4
#define MAX_NUM_OF_TUNNEL           3   /* Num of Tunnel per Bearer */
#define MAX_NUM_OF_PF               16  /* Num of Packet Filter per Bearer */

#define MAX_POOL_OF_UE              (MAX_NUM_OF_ENB * MAX_NUM_OF_UE)
#define MAX_POOL_OF_SESS            (MAX_POOL_OF_UE * MAX_NUM_OF_SESS)
#define MAX_POOL_OF_BEARER          (MAX_POOL_OF_SESS * MAX_NUM_OF_BEARER)
#define MAX_POOL_OF_TUNNEL          (MAX_POOL_OF_BEARER * MAX_NUM_OF_TUNNEL)
#define MAX_POOL_OF_PF              (MAX_POOL_OF_BEARER * MAX_NUM_OF_PF)
#define MAX_POOL_OF_DIAMETER_SESS   (MAX_POOL_OF_UE * MAX_NUM_OF_SESS)

Currently, the number of UE is limited to 128*128.

What is the Default Configuration?

Network

* MME
  S1AP: listen on all address avaiable in system
  GTP-C: listen on the first IP address in system
  DIAMETER: 127.0.0.2 (No TLS)

* SGW
  GTP-C: 127.0.0.2
  GTP-U: listen on the first IP address in system

* PGW
  GTP-C: Both 127.0.0.3 and [::1]
  GTP-U: Both 127.0.0.3 and [::1]
  DIAMETER: 127.0.0.3 (No TLS)

* HSS
  DIAMETER: 127.0.0.4 (No TLS)

* PCRF
  DIAMETER: 127.0.0.5 (No TLS)

GUMMEI, PLMN and TAC

* GUMMEI
  PLMN ID - MNC: 001, MCC: 01
  MME Group : 2
  MME Code : 1

* TAI
  PLMN ID - MNC: 001, MCC: 01
  TAC : 12345

Security

* Integrity : EIA1 - Snow 3G
* Ciphering : EEA0 - Nothing

UE Network

* IPv4 : 45.45.0.1/16
* IPv6 : cafe::1/64

* IPv4
  Primary : 8.8.8.8 
  Secondary : 8.8.4.4

* IPv6
  Primary : 2001:4860:4860::8888
  Secondary : 2001:4860:4860::8844

The parsing errors are caused by tab spaces in the configuration files.

YAML forbids tabs. You should use space instead of tab in NextEPC configuration file.

YAML FAQ: Why does YAML forbid tabs?

Tabs have been outlawed since they are treated differently by different editors and tools. And since indentation is so critical to proper interpretation of YAML, this issue is just too tricky to even attempt. Indeed Guido van Rossum of Python has acknowledged that allowing TABs in Python source is a headache for many people and that were he to design Python again, he would forbid them.

Cross compilation setup

By default, NextEPC is designed to support the Embedding System. To do so, we introduced pool-based memory management. Unfortunately, we have not tested NextEPC in an embedded environment. I tried to compile on the ARM architecture using Docker and run it with QEMU for your reference.

We'll use Debian Docker Environment.

$ git clone -r https://github.com/acetcom/nextepc
$ cd nextepc/docker
$ DIST=debian docker-compose run dev

In Docker Container

acetcom@nextepc-dev:~$ sudo dpkg --add-architecture armel
acetcom@nextepc-dev:~$ sudo apt-get install libsctp-dev:armel libyaml-dev:armel libgnutls28-dev:armel libgcrypt-dev:armel libssl-dev:armel libmongoc-dev:armel libbson-dev:armel
acetcom@nextepc-dev:~$ sudo apt-get install crossbuild-essential-armel
acetcom@nextepc-dev:~$ sudo apt-get install qemu

acetcom@nextepc-dev:~$ mkdir git
acetcom@nextepc-dev:~$ cd git/
acetcom@nextepc-dev:~/git$ git clone https://github.com/acetcom/nextepc
acetcom@nextepc-dev:~/git$ cd nextepc/
acetcom@nextepc-dev:~/git/nextepc$ autoreconf -if;./configure --prefix=`pwd`/install --host=arm-linux-gnueabi;make -j 2
acetcom@nextepc-dev:~/git/nextepc$ make install
acetcom@nextepc-dev:~/git/nextepc$ qemu-arm .libs/nextepc-mmed
NextEPC daemon v0.3.10 - Oct  4 2018 13:24:24

[10/04 13:38:06.329] WARN: pid file /home/acetcom/git/nextepc/install/var/run/nextepc-mmed/pid overwritten -- Unclean shutdown of previous NextEPC run? (application.c:113)
  PID[55780] : '/home/acetcom/git/nextepc/install/var/run/nextepc-mmed/pid'
  File Logging : '/home/acetcom/git/nextepc/install/var/log/nextepc/nextepc.log'
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
qemu: Unsupported syscall: 345
  MongoDB URI : 'mongodb://mongodb/nextepc'
  Configuration : '/home/acetcom/git/nextepc/install/etc/nextepc/nextepc.conf'
[10/04 13:38:06.400] MME try to initialize
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown QEMU_IFLA_INFO_KIND ipip
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown QEMU_IFLA_INFO_KIND ip6tnl
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown QEMU_IFLA_INFO_KIND ipip
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown QEMU_IFLA_INFO_KIND ip6tnl
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
Unknown host QEMU_IFLA type: 40
Unknown host QEMU_IFLA type: 41
[10/04 13:38:08.693] gtp_server() [172.20.0.2]:2123
[10/04 13:38:08.696] gtp_client() [127.0.0.2]:2123
Unsupported setsockopt level=132 optname=11
[10/04 13:38:08.697] ERRR: Unable to subscribe to SCTP events: (92:Protocol not available) (unix/sctp.c:291)
[10/04 13:38:08.699] ASSERT: !(rv == CORE_OK).  (unix/sctp.c:33)
[10/04 13:38:08.700] ERRR: sctp_server() [172.20.0.2]:36412 failed(92:Protocol not available) (unix/sctp.c:98)
[10/04 13:38:08.701] ASSERT: !(rv == CORE_OK).  (s1ap_sctp.c:35)
[10/04 13:38:08.702] ASSERT: !(rv == CORE_OK).  (s1ap_path.c:53)
[10/04 13:38:08.703] ASSERT: !(rv == CORE_OK).  (s1ap_path.c:28)
[10/04 13:38:08.704] ERRR: Can't establish S1AP path (mme_sm.c:63)
[10/04 13:38:08.708] MME initialize...done


[10/04 13:38:08.710] INFO: NextEPC daemon start (main.c:157)

The SCTP module is not included in the QEMU kernel. I believe that if the Linux kernel installed on your target platform contains an SCTP module, it will work normally. {: .notice--warning}

16 KiB Raw Blame History