gmane.comp.mobile.openbsc.general

Re: SimpleHLR: A web interface for OpenBSC

Alton MacDonald — 2013-05-17T15:57:32

Hi,

We were using this interface for a while on our GSM testbed.
Unfortunately, after having many subscribers, the web interface would
not load correctly and would suffer from performance issues.

Since I had also done a few python scripts to help manage our local
OpenBSC installation, I used your interface as inspiration to design my
own and incorporate my python scripts as well.

Which brings me to proposing another web interface for OpenBSC and
OsmoSGSN. I've attached a README file with the additional features my
interface provides.

On one hand, it adds some functionality to OpenBSC which sends welcome
messages to subscribers the first time they are seen on the network
and/or return from roaming.

On the other hand, it facilitates the administration of MANY subscribers
by providing a filterable interface where authorization can be quickly
toggled. It also allows for changing the timeslot configurations so that
an operator can adapt his/her network to the current required Voice and
Data traffic.

Since I

Siemens BS-11 available

Norbert Hüttisch — 2013-05-19T19:57:08

I've got a Siemens BS-11 available. For free. It has a 220/230V power supply and 2
GSM900 TRX.

It must be picked up in Karlsruhe.

If interested, please mail.

N.

Re: Testing Openbsc without a BTS

Holger Hans Peter Freyther — 2013-05-16T11:50:34


the mail was wrong. One needs to create a file in this directory and any
name will work out. I have created this[1] wiki page and corrected the
information from the mail.

holger

[1] http://openbsc.osmocom.org/trac/wiki/simulation

Re: New handover algorithm

jolly — 2013-05-16T11:10:46

good advice!

http://openbsc.osmocom.org/trac/wiki/handover

Re: New handover algorithm

Harald Welte — 2013-05-16T10:44:10

Hi Jolly,

thanks a lot for working on this.

As a side-note: I think there is little point in referring to a PDF file
on a web server, especially if that PDF file only contains plain text.
Either send that full plain text to this mailing list for discussion,
and/or maintain the text in the wiki.  This way it can be maintained
collaboratively and stays with the other information related to the
project.

Regards,
Harald

New handover algorithm

jolly — 2013-05-16T08:01:15

Hi,

Jean-Samuel and me developed an advanced handover algorithm for OpenBSC.
Before implementing it, we would like to get any feedback, so I decided
to post it in this mailing list. The document that describes the
algorithm can be downloaded at:
http://home.eversberg.eu/handover-algorithm-v2.3.pdf

Regards,

Andreas



Handover shall ensure that the mobile station is always on the best cell
during a call. This results in best speech quality and allows a lower TX
level (up-link) to save battery and reduce interference. Sometimes this
is not desirable, especially if a cell with large coverage becomes
congested after handover. In this situation it makes sense to balance
the usage between cells to prevent congestion. If a cell is congested,
one or several mobile stations need to be handovered to a different
cell, even if that cell is not the best cell for these mobile station.
In case that a mobile station needs to handover to prevent a loss of
radio link, it makes sense to handover it to a cell that is congest

Re: Testing Openbsc without a BTS

Priyanka Behl — 2013-05-14T06:24:47

Hi,

I am sorry but I got this Failure notice and that's why i re-send it. But I'll take care of this in future.

________________________________
 From: Holger Hans Peter Freyther <holger< at >freyther.de>
To: Priyanka Behl <priyankabehl< at >ymail.com> 
Cc: "openbsc< at >lists.osmocom.org" <openbsc< at >lists.osmocom.org> 
Sent: Tuesday, 14 May 2013 11:43 AM
Subject: Re: Testing Openbsc without a BTS

On Tue, May 14, 2013 at 12:53:17PM +0800, Priyanka Behl wrote:

Hi,

your behavior of re-sending a mail within a day is considered pushy
in our community!

Just think about the error message and the above command and you will
find a solution.

Re: Testing Openbsc without a BTS

Holger Hans Peter Freyther — 2013-05-14T06:13:59

Hi,

your behavior of re-sending a mail within a day is considered pushy
in our community!


Just think about the error message and the above command and you will
find a solution.

Re: Testing Openbsc without a BTS

Priyanka Behl — 2013-05-14T04:53:17

Hi, As per your mails about a pure software BTS when I tried installing it as:
root< at >server:~# echo 'deb http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./' > /etc/apt/sources.list.d/

I got the following error
bash:  /etc/apt/sources.list.d/: Is a directory
so next i tried installing it as:
root< at >server:~# echo 'deb http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./' > /etc/apt/sources.list
root< at >server:~# apt-get update
Get:1 http://download.opensuse.org/ ./ Release.gpg [189 B]
Ign http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./ Translation-en
Ign http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./ Translation-en_IN
Get:2 http://download.opensuse.org/ ./ Release [463 B]
Ign http://download.opensuse.org/ ./ Release
Ign http://download.opensuse.org/ ./ Packages/DiffIndex
Ign http://download.opensuse.org/ ./ Packages
Hit http://download.opensuse.org/ ./ Packages
Fetched 190 B in 1s (145 B/s)
Reading package lists... Done
W: GPG err

Testing Openbsc with software BTS

Priyanka Behl — 2013-05-14T04:42:05

Hi,
 As per your mails about a pure software BTS when I tried installing it as:
root< at >server:~# echo 'deb http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./' > /etc/apt/sources.list.d/

I got the following error
bash:  /etc/apt/sources.list.d/: Is a directory
so next i tried installing it as:
root< at >server:~# echo 'deb http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./' > /etc/apt/sources.list
root< at >server:~# apt-get update
Get:1 http://download.opensuse.org/ ./ Release.gpg [189 B]
Ign http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./ Translation-en
Ign http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./ Translation-en_IN
Get:2 http://download.opensuse.org/ ./ Release [463 B]
Ign http://download.opensuse.org/ ./ Release
Ign http://download.opensuse.org/ ./ Packages/DiffIndex
Ign http://download.opensuse.org/ ./ Packages
Hit http://download.opensuse.org/ ./ Packages
Fetched 190 B in 1s (145 B/s)
Reading package lists... Done
W: GPG e

Re: Testing Openbsc without a BTS

Priyanka Behl — 2013-05-13T07:58:58

Hi,
 As per your mails abut a pure software BTS when I tried installing it as:
root< at >server:~# echo 'deb http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./' > /etc/apt/sources.list.d/

I got the following error
bash:  /etc/apt/sources.list.d/: Is a directory

so next i tried installing it as:

root< at >server:~# echo 'deb http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./' > /etc/apt/sources.list
root< at >server:~# apt-get update
Get:1 http://download.opensuse.org/ ./ Release.gpg [189 B]
Ign http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./ Translation-en
Ign http://download.opensuse.org/repositories/home:/zecke23/Debian_6.0/ ./ Translation-en_IN
Get:2 http://download.opensuse.org/ ./ Release [463 B]
Ign http://download.opensuse.org/ ./ Release
Ign http://download.opensuse.org/ ./ Packages/DiffIndex
Ign http://download.opensuse.org/ ./ Packages
Hit http://download.opensuse.org/ ./ Packages
Fetched 190 B in 1s (145 B/s)
Reading package lists... Done
W: GPG

Re: [PATCH] nat: fix use after free in forward_sccp_to_bts

Holger Hans Peter Freyther — 2013-05-12T20:13:21

interesting that it never showed up when I used valgrind and the code
appears to be like this since 2010. Feel free to push this to master.

thanks
holger

[PATCH] nat: fix use after free in forward_sccp_to_bts

2013-05-12T18:52:15

From: Pablo Neira Ayuso <pablo< at >gnumonks.org>

valgrind detected an use after free in the path of forward_sccp_to_bts.
The 'parsed' object is referenced from update_con_authorize.
---
 openbsc/src/osmo-bsc_nat/bsc_nat.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/openbsc/src/osmo-bsc_nat/bsc_nat.c b/openbsc/src/osmo-bsc_nat/bsc_nat.c
index b9bf36c..be8d56a 100644
--- a/openbsc/src/osmo-bsc_nat/bsc_nat.c
+++ b/openbsc/src/osmo-bsc_nat/bsc_nat.c
< at >< at > -718,15 +718,18 < at >< at > static int forward_sccp_to_bts(struct bsc_msc_connection *msc_con, struct msgb *
 LOGP(DNAT, LOGL_ERROR, "Unknown connection for msg type: 0x%x from the MSC.\n", parsed->sccp_type);
 }
 
-talloc_free(parsed);
-if (!con)
+if (!con) {
+talloc_free(parsed);
 return -1;
+}
 if (!con->bsc->authenticated) {
+talloc_free(parsed);
 LOGP(DNAT, LOGL_ERROR, "Selected BSC not authenticated.\n");
 return -1;
 }
 
 update_con_authorize(con, parsed, msg);
+talloc_free(parsed);
 
 bsc_send_data(con->bsc, msg

Re: Testing Openbsc without a BTS

Holger Hans Peter Freyther — 2013-05-09T08:29:25

Hi,


you might want to look at my mails about a pure software BTS used to
test OpenBSC. The mails are here[1][2].



This is to be expected. Our system enforces a periodic location
updating procedure and if a subscriber will not do that, it will
be expired. Your subscriber needs to make a Location Updating Procedure.

holger


[1] http://lists.osmocom.org/pipermail/openbsc/2012-October/004210.html
[2] http://lists.osmocom.org/pipermail/openbsc/2012-October/004250.html

Testing Openbsc without a BTS

Priyanka Behl — 2013-05-09T06:24:55

Hi everyone,
I am trying to test Openbsc without a BTS. I need your help as I am very new to Openbsc.
After doing the configurations for openbsc-multitrx.cfg file this is how I started the process:
cd openbsc/openbsc
src/osmo-nitb -c doc/examples/osmo-nitb/nanobts/openbsc-multitrx.cfg -C


And the output I received was:
<0019> input/ipaccess.c:925 enabling ipaccess BSC mode
DB: Database initialized.
DB: Database prepared.
<0001d> sms_queue.c:220 Attempting to send 20 SMS


After this I added a subscriber in the hlr and the subscriber was added in the output:
 <0019> ipaccess.c:841 enabling ipaccess BSC mode
DB: Database initialized.
DB: Database prepared.
<001d> sms_queue.c:232 Attempting to send 20 SMS
<0002> gsm_subsciber.c:390 Expiring inactive subscriber Testing(ID 1)

After this I used the following commands on the osmo-nitb console and receive the following output:
Openbsc>show subscrber id 1
            ID:  1,    Authorized:  1
            Name:    'Testing'

Re: Silent Call vs Test Call vs A-bis Proxy (ip.access-proxy)

Holger Hans Peter Freyther — 2013-05-08T17:52:59

okay, the only free support I will give is to improve the wiki. If
you have questions regarding specific pages or proposals for page
I will consider to clarify/create.

holger

Re: Silent Call vs Test Call vs A-bis Proxy (ip.access-proxy)

j h — 2013-05-08T17:43:51

I do not have an account. I've been looking through the OpenBSC
docs/codebase, and a few of Harald's presentations.

On Wed, May 8, 2013 at 12:51 PM, Holger Hans Peter Freyther
<holger< at >freyther.de> wrote:

Re: Silent Call vs Test Call vs A-bis Proxy (ip.access-proxy)

Holger Hans Peter Freyther — 2013-05-08T16:51:58

Hi,

do you have an account for our wiki? Where did you search for the
information in it?

holger

Silent Call vs Test Call vs A-bis Proxy (ip.access-proxy)

j h — 2013-05-08T13:35:27

Could someone explain to me the difference between Silent Call, Test Call,
and A-bis proxy (ip.access-proxy).
To my understanding Test Call opens a pass through UDP port straight to the
MS. While SilentCall just creates just creates a open channel.

My confusion I believe comes from where you're "sitting" in the eco-system
between Silent Call and Test Call. I also am confused after looking at
ip.access-proxy. It clearly opens a port for injection either towards the
BSC or BTS. However, does a silent-call need to be active to get data to
the MS? When I try to inject a simple SMS using ip.access-proxy, (via
copying traffic seen across ip.access), I get length disagrees with header.
What is ip.access injection expecting in terms of data format?

Thanks

Re: Pysim error: SW match failed

Lukas Kuzmiak — 2013-05-08T10:47:36

btw, you may want to move this discussion to the simtrace mailing list
instead of openbsc, I think it fits there better.

Lukas


On Wed, May 8, 2013 at 1:46 PM, Lukas Kuzmiak <lukash< at >backstep.net> wrote:

Re: Pysim error: SW match failed

Lukas Kuzmiak — 2013-05-08T10:46:36

Hey there,

it seems to me the pySim was trying to update 6f30 file (not sure why):
  self._scc.update_binary('6f30', hplmn + 'ff' * (tl-3))

and got back 0x9804 instead of 0x9000, which means you don't have
sufficient permissions to update that file - either pin verification went
wrong or you need some other key to update that file, not sure what
permissions are set of 6f30 file there.

Hopes this will help you debugging it.

Cheers!
Lukas


On Wed, May 8, 2013 at 1:38 PM, Martin Nagy <nugett< at >centrum.sk> wrote: