VOLTE on Openstack

[ghost]

Dear @herlesupreeth Hi again

I managed to run IMS+Core on Openstack with internal ip 10.4.128.21/24 and Floating Ip address of 10.20.20.177 which both of them are on a Single PC with 192.168.100.20 ip addresses. and installed SRSLTE on the other PC with ip address of 192.168.100.2 so now eNB is connected to MME and UE registered on the network. But i do not have internet access! I have done these to be able to connect eNB to Core : sudo route add -net 10.20.20.0/24 gw 192.168.100.20 enp1s0 and Now i tried your command: ip r add 10.4.128.21/24 via 10.20.20.177 but it says:

ERROR: INVALID prefix for given prefix length.

and I dont have internet access!!

and Also Tested SRSEPC commands sudo srsepc_if_masq enp1s0

But still not access!

also in sgw:

sgw:
    gtpc:
      addr: 127.0.0.2
    gtpu:
      dev: ens3
      advertise_addr: 10.20.20.177

i had set advertise_addr! Can u please give me a hint?

Thanks Best regards

P.S. I set UE's pool in pgw.yaml

 ue_pool:
      - addr: 172.20.110.1/24
        dev: ogstun
        apn: internet
      - addr: 172.20.111.1/24
        apn: ims
        dev: ogstun2
    dns:
      - 8.8.8.8
      - 8.8.4.4
      - 2001:4860:4860::8888
      - 2001:4860:4860::8844
    p-cscf:
      - 10.4.128.21

[ghost]

Is there sth wrong with UE ip addresses? or any where to add route?

[herlesupreeth]

is this 10.20.20.177 publicly reachable range? make sure eNB is in the same range as Floating IP Address

[ghost]

this IP address is pingable from eNB and actually eNB is Connected to MME but ue does not have internet access

[ghost]

when i compiled OpenStack this was the default External Ip address! (10.20.20.0) but my whole network is in range of 192.168.100.0 network

i used command in eNB to be able to route to MME sudo route add -net 10.20.20.0/24 gw 192.168.100.20 enp1s0

[ghost]

here is my openstack images

[herlesupreeth]

this IP address is pingable from eNB and actually eNB is Connected to MME but ue does not have internet access

In sgwu.yaml configuration file make sure to add the following under gtpu section

...
    gtpu:
      - addr: 10.4.128.21
        advertise_addr: 10.20.20.177
...

[ghost]

I'm using 4G network with open5gs v1.3 i have sgw

As i mentioned earlier

also in sgw:

sgw:
    gtpc:
      addr: 127.0.0.2
    gtpu:
      dev: ens3
      advertise_addr: 10.20.20.177

i set in this way do i need to add this?

• addr: 10.4.128.21 advertise_addr: 10.20.20.177 ...

[ghost]

I did what u said but still do not have internet access

When i Turn Volte On in UE i can see some green lines which shows it is trying to connect but pgw shows this 02/23 14:32:32.909: [pgw] WARNING: [DROP] Cannot find PGW S5U bearer : TEID[0x7] (../src/pgw/pgw-gtp-path.c:180)

[herlesupreeth]

I suspect your OpenStack networking setup is wrong. Can you make sure you have opened all the tcp and udp ports in the Security Policies in OpenStack for that VM?

[ghost]

I made a mistake on Startup bash so now i have very poooor internet connection! and Volte is try to connect but cant!

I can not open Google in browser but could see a clip in whats app!!!!

[ghost]

my openstack pc is 6.6% utility of cpu and 20% RAM!

[ghost]

So weirdo!

[herlesupreeth]

Make sure to lower the MTU of ogstun and ogstun2 to 1400 as the ethernet interface on OpenStack VMs have an MTU of 1450

[ghost]

same thing happens! but sometimes do not have access to internet at all!

[herlesupreeth]

Sorry cant help.. please debug your setup as I suspect its at fault here

[ghost]

Thanks anyway🙏🙏🙏

[ghost]

Dear @herlesupreeth

Sorry to bother you again, I removed everything and start again from scratch

Now i have stable internet access but couldn't register on VOLTE!

here is the trace: VOLTE not register on openstack.zip

[ghost]

Hi The following is full pcap from running components to ue attachment

Could u please check this? Not_registered.zip

[ghost]

I also have these errors on pcrf

02/25 07:26:36.021: [diam] ERROR: ../subprojects/freeDiameter/libfdcore/p_psm.c:287 'STATE_OPEN' -> 'STATE_CLOSED' 'pcscf.ims.mnc001.mcc001.3gppnetwork.org' (../lib/diameter/common/init.c:116)

02/25 06:59:30.312: [diam] ERROR: ../subprojects/freeDiameter/libfdcore/p_psm.c:630 ERROR: in '(fd_p_ce_msgrcv(&msg, (hdr->msg_flags & 0x80), peer))' : Invalid argument (../lib/diameter/common/init.c:116)

Is there sth wrong with freediameter?

[herlesupreeth]

There is something wrong with the connection between fhoss and icscf, in the pcap it shows that UAR is not replied by fhoss..Check the connection there

[ghost]

Thanks

Dear @herlesupreeth

I run that again from scratch! i have some errors when want to run PCRF! I think the problem is also related to this!

is this p_psm.c the problem?

[ghost]

when

i run pcscf,scscf,icscf and also all EPC components there is noting wrong , just not connect to FHOSS, and all goes well as u can see in bellow images:

icscf:

pcscf:

When

i run FHOSS and when i connect single UE without Volte enabled images are as bellow:

icscf:

pcscf:

At the END when

i ENABLE VOLTE on UE images are as bellow:

iscsf:

pcscf:

FHOSS

PCRF:

It means when i run all components there is noting wrong ! But when The UE tries to Enable VOLTE it gets stuck in there!!!!

Can u please check if the above message is related or not?

p.s: I'm using open5gs V1.3 which as only EPC and components are as follows pgw ,sgw ,hss ,mme ,pcrf, all CORE component are in a single VM on OPENSTACK

Still i think there is sth wrong with PCRF!

[ghost]

While installing
when i enter this command: dpkg -i *.deb I face with an error:

I think its usual and after setting RTP will go!

[herlesupreeth]

I think its usual and after setting RTP will go!

That's right. Its clearly written in the tutorial that once you copy the required config file and restart this error will go away