From nhofmeyr at sysmocom.de  Thu Nov  1 00:29:16 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Thu, 1 Nov 2018 01:29:16 +0100
Subject: [osmocom-event-orga] 11.11.: GSM workshop + spectrum/hardware
 discussion @ AfRA
Message-ID: <20181101002916.GB4630@my.box>

A reminder that on Sunday, 11th of November, I will host an informal "How does
Osmocom GSM / UMTS work" workshop at the AfRA, 15-19h.

I would also like to discuss the situation about the lack of hardware-spectrum-
match, which is for me the most urgent topic right now. If we don't have
anything to run on, we might as well give the room to someone else.

So please come in person, as many of the GSM team as possible, so we can
freely brainstorm in private.

If anyone has any ideas in the meantime to get a hardware-spectrum match,
please pursue and let us know!

We can do the discussion ahead of the workshop, or as an intermission around
the middle, or at the end. I don't mind, and would decide on that depending on
how many of you are present at what time. Feedback welcome!

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181101/5a83676d/attachment.bin>

From ml at mail.tsaitgaist.info  Thu Nov  1 14:59:34 2018
From: ml at mail.tsaitgaist.info (=?iso-8859-1?Q?K=E9vin?= Redon)
Date: Thu, 1 Nov 2018 15:59:34 +0100
Subject: [osmocom-event-orga] Nano3G-E16 Re: No time for 35c3 gsm after
 all
In-Reply-To: <20181031221740.GJ27112@nataraja>
References: <20181028075711.GE12047@ted.stsp.name>
 <20181030073428.GS31654@nataraja> <20181031221740.GJ27112@nataraja>
Message-ID: <20181101145934.GA2590@coil>

On Wed, Oct 31, 2018 at 11:17:40PM +0100, Harald Welte wrote:
> My opinion is that unless somebody with significant knowledge on femtocell hacking
> (Kevin?) has time to work on this, the nano3G-E16 are not going to in any shape
> where they can be used at an event, sorry.

I will be back in Berlin next week and can have a look at it.


From arne at nerdkeller.org  Thu Nov  1 20:51:13 2018
From: arne at nerdkeller.org (Arne Zachlod)
Date: Thu, 1 Nov 2018 21:51:13 +0100
Subject: [osmocom-event-orga] 11.11.: GSM workshop + spectrum/hardware
 discussion @ AfRA
In-Reply-To: <20181101002916.GB4630@my.box>
References: <20181101002916.GB4630@my.box>
Message-ID: <ae24361a-e179-6a21-6380-d3093e8698ce@nerdkeller.org>

Hi all,

just some not very creative idea, but why don't we ask the orga if we 
could invest the funds to buy the BTSs from sysmocom? Is it just the 
question of getting the funds or is there some other factor?

- Arne

On 11/1/18 1:29 AM, Neels Hofmeyr wrote:
> A reminder that on Sunday, 11th of November, I will host an informal "How does
> Osmocom GSM / UMTS work" workshop at the AfRA, 15-19h.
> 
> I would also like to discuss the situation about the lack of hardware-spectrum-
> match, which is for me the most urgent topic right now. If we don't have
> anything to run on, we might as well give the room to someone else.
> 
> So please come in person, as many of the GSM team as possible, so we can
> freely brainstorm in private.
> 
> If anyone has any ideas in the meantime to get a hardware-spectrum match,
> please pursue and let us know!
> 
> We can do the discussion ahead of the workshop, or as an intermission around
> the middle, or at the end. I don't mind, and would decide on that depending on
> how many of you are present at what time. Feedback welcome!
> 
> ~N
> 


From lynxis at fe80.eu  Thu Nov  1 22:04:48 2018
From: lynxis at fe80.eu (Alexander Couzens)
Date: Thu, 1 Nov 2018 23:04:48 +0100
Subject: [osmocom-event-orga] No time for 35c3 gsm after all
In-Reply-To: <20181028075711.GE12047@ted.stsp.name>
References: <20181028075711.GE12047@ted.stsp.name>
Message-ID: <20181101230448.106a9d0a@lazus.yip>

On Sun, 28 Oct 2018 08:57:11 +0100
Stefan Sperling <stsp at stsp.name> wrote:

> I am sorry about this, but I realised that I won't have time to help
> out with the GSM team after all.

Thanks for your help!

Best,
lynxis
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181101/da591612/attachment.bin>

From lynxis at fe80.eu  Thu Nov  1 22:19:38 2018
From: lynxis at fe80.eu (Alexander Couzens)
Date: Thu, 1 Nov 2018 23:19:38 +0100
Subject: [osmocom-event-orga] update from POC
In-Reply-To: <20181027182046.GB14737@my.box>
References: <20181027182046.GB14737@my.box>
Message-ID: <20181101231938.62548c34@lazus.yip>

Thanks for the update!

> * POC offers to locate the prod core in a VM on their machine

I would prefer using our own machine.
 
> * a POC core machine is set up and running already now, before
> congress
> * neels will get VPN access (we can test against that)

nice. so we can test everything in advance.

> * POC now use several Yate instances, one per realm (DECT, SIP,
> GSM,...)

I also would stay at freeswitch to have more control over the
connection to yate. It might be a lot easier, because freeswitch can do
transcoding and force certain codecs and we can debug it easier, it's
then clear, the bug is on our side or not.

> POC suggests:
> * We can ask Lime to send us the LimeSDR devices used on EFR camp
> (ask POC, via neels) to use them as GSM BTS

I would like to use sysmobts, if possible for 2g.
 
> * Idea: share one secure network segment with the POC

Our gsm server can have a another leg in the poc network, so we don't
walk over the public network with the voice traffic. But for the
basestation I would like to be on our vlan (especially the femtocells).

Best,
lynxis
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181101/0a1a343a/attachment.bin>

From nhofmeyr at sysmocom.de  Sun Nov  4 21:23:38 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Sun, 4 Nov 2018 22:23:38 +0100
Subject: [osmocom-event-orga] next sunday
Message-ID: <20181104212338.GB22559@my.box>

Hi all,

Re: frequencies and HW for the Congress: It seems that I have misunderstood a
bit, it's not quite as final as I thought.

a) it's fairly certain but not 100% yet that we will get the 850 band ARFCN.
b) the EU band is not final yet, we might still get one with some luck.
c) even if only on 850, maybe we can still get 2G hardware.

We'll go on gearing up for the congress as if we had ARFCNs+BTS and will likely
do *something* either way.


can I get a quick wave of people planning to attend next sunday, 11.11., 15h at
AfRA?

_o/

If you're coming, are you coming to learn about Osmocom GSM, or for discussion?
I plan to do a bit of both, in which order is yet to be determined.

Thanks!

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181104/d088ae09/attachment.bin>

From ralph at schmid.xxx  Tue Nov  6 15:06:04 2018
From: ralph at schmid.xxx (Ralph A. Schmid, dk5ras)
Date: Tue, 6 Nov 2018 16:06:04 +0100
Subject: [osmocom-event-orga] update from POC
In-Reply-To: <20181029125106.GA4028@my.box>
References: <20181027182046.GB14737@my.box>
 <CAPEnp=bpx7cZtraWSrPFFt5rPm=+hiJRdu9i6ScDgne1DrV5Qw@mail.gmail.com>
 <20181029125106.GA4028@my.box>
Message-ID: <002701d475e2$3d187a20$b7496e60$@schmid.xxx>

Hi,

> But my general feeling with all SDR solutions is "too flaky to be fun", so
I
> won't push in that direction.

My experiences with SDRs are quite positive. Especially small cells covering
only one or two rooms work quite good, even without duplexer, and only with
two rabbit ear TX/RX antennae. Even a RasPi 3B is powerful enough to run a
nice cell with two ARFCNs for voice and GPRS and a BladeRF SDR as TRX. 

With the LimeSDR I only played with narrow band stuff like osomocom-analog,
but this also runs very stable on a RasPi, delivering stable analog cell
coverage over days without any issues. Should be not so different with GSM.
 
About this 850 MHz assignment, what ARFCNs are officially available? I do my
private tests (carefully) on license free channels, as close as possible to
the SRD regulations for downlink (for example 869.8 MHz or so), and after
all the uplink anyway is somehow a wireless mic, isn't it?! :)))

> ~N

Ralph.




From nhofmeyr at sysmocom.de  Tue Nov  6 19:50:52 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Tue, 6 Nov 2018 20:50:52 +0100
Subject: [osmocom-event-orga] update from POC
In-Reply-To: <002701d475e2$3d187a20$b7496e60$@schmid.xxx>
References: <20181027182046.GB14737@my.box>
 <CAPEnp=bpx7cZtraWSrPFFt5rPm=+hiJRdu9i6ScDgne1DrV5Qw@mail.gmail.com>
 <20181029125106.GA4028@my.box>
 <002701d475e2$3d187a20$b7496e60$@schmid.xxx>
Message-ID: <20181106195052.GC19568@my.box>

On Tue, Nov 06, 2018 at 04:06:04PM +0100, Ralph A. Schmid, dk5ras wrote:
> My experiences with SDRs are quite positive.

But for a "production" setup, you do get all sorts of odd errors showing up
that make you examine potential bugs and whatnot, when they were simply timing
errors from the lack of SDR timer precision.

The Ettus B210 worked pretty well for me, at least for some while. Every now
and then nonsense came up. At some points I wasn't able to attach for many
attempts in a row, etc. You can add a high precision timing source or a GPSDO
to alleviate that. But in the end a proper sysmoBTS cuts short all of that BS
and gives me the best fun per effort ratio.

It happens more often than not that we get bug reports from customers for some
Osmocom misbehavior, and it turns out the cause was an SDR with improper
timing accuracy.

Especially with testing handover, I found that even two sysmoBTS fail when they
are improperly calibrated. I don't think an SDR without GPSDO/external clock
has even a chance there. (The XTRX claims to have good timing, haven't tested)

> Especially small cells covering only one or two rooms work quite good, even
> without duplexer, and only with two rabbit ear TX/RX antennae.

To create coverage for 35c3? :)

> Even a RasPi 3B is powerful enough

The core net system isn't that critical, you can even run Abis over Satellite
backhaul with near infinite latency. The problem is tightness of the RF.

> With the LimeSDR I only played with narrow band stuff like osomocom-analog,
> but this also runs very stable on a RasPi, delivering stable analog cell
> coverage over days without any issues. Should be not so different with GSM.

Let's not make assumptions on GSM based on analog cell :)
I have no idea about analog cell, but I assume it is worlds apart...

> About this 850 MHz assignment, what ARFCNs are officially available?

So far none yet, will probably keep you posted here.

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181106/c93d33e2/attachment.bin>

From nhofmeyr at sysmocom.de  Tue Nov 13 16:10:11 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Tue, 13 Nov 2018 17:10:11 +0100
Subject: [osmocom-event-orga] good news everyone: BTS and bands
Message-ID: <20181113160428.GA18310@my.box>

So, today I rummaged throught the event pool boxes at sysmocom office.

Turns out:

a) I can find only two BTS in there. There are four amplifier+duplexer wooden
boards, but only two of them have a BTS on them. Does anyone know where the
other two went?

b) The good news is: both BTSes that I tested are quad band, i.e. they support
whichever band we want. Lynxis and I also each have one quad-band sysmoBTS. So
we have at least four BTS available now, I assume it will be six in the end.
(Details: a sysmoBTS is physically limited to specific bands. Which bands are
supported is indicated in the EEPROM, and if you configure osmo-bts.cfg with a
band that isn't supported, you get, during osmo-bts startup:
"DL1C <0006> l1_if.c:1453 BTS band GSM850 not supported by hw"
i.e. the two that worked at AfRA on sunday positively support the 850 band.)

c) To transceive with 2W, we need amps and duplexers.

The current four amps in the event pool are 1800; sysmocom has more than enough
lo-band amp boards that we could replace in these amps with "minimal mechanical
effort".

Sysmocom has three 850 duplexers available that we can transfer to the event
pool.


In summary:

If we get 850MHz spectrum:
We are limited by duplexers to three sysmoBTS 850MHz on 2W, and could setup
another one to three sysmoBTS 850MHz on 200mW.

If we get 1800MHz spectrum:
We can have four BTS on 2W like last year, and could setup two more on 200mW.
There are also two 1800 nanoBTS in the pool.

This is pretty good news, it means that whichever spectrum we get, we can do 2G
with at least three 2W cells. Now let's wait/hope for at least one license
confirmation, I think that is quite likely to happen.

Maybe we can even get both 850 and 1800, then we could do 2G on 1800 and 3G on
850. There are eight nano3G in the event pool, plus some more from me, lynxis,
maybe others.

~N


From nhofmeyr at sysmocom.de  Tue Nov 13 16:19:17 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Tue, 13 Nov 2018 17:19:17 +0100
Subject: [osmocom-event-orga] 3G voice: IuUP
Message-ID: <20181113161917.GA20168@my.box>

I've been asked about IuUP specs.

So far I know of 3GPP TS 25.414, 5.1.3.3.2 RTP Payload:
"A single Iu UP PDU, as described in TS 25.415 [24], shall be transported as RTP payload."

And 25.415 contains the juicy bits.

We have osmo-mgw branch neels/iuup refactoring some RTP packet handling to be
able to easily strip/insert an IuUP header with correct checksums etc., and we
have libosmocore branch laforge/iu_up implementing a state machine about
unshuffling the payload bits. I guess to get a working solution we would need
to marry those two branches and fix whatever is still broken/not implemented in
there. They are still potentially wildly incompatible.

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181113/dee10ead/attachment.bin>

From frank at ccc.de  Tue Nov 13 16:23:21 2018
From: frank at ccc.de (Frank Rieger)
Date: Tue, 13 Nov 2018 17:23:21 +0100
Subject: [osmocom-event-orga] good news everyone: BTS and bands
In-Reply-To: <20181113160428.GA18310@my.box>
References: <20181113160428.GA18310@my.box>
Message-ID: <0C51C82F-5B1A-4812-9B65-47D7A10FBAAD@ccc.de>

I have no word from Deutsche Telekom so far. Has anyone else progress with other operators or the BNetzA?

Frank

---

> On 13. Nov 2018, at 17:10, Neels Hofmeyr <nhofmeyr at sysmocom.de> wrote:
> 
> So, today I rummaged throught the event pool boxes at sysmocom office.
> 
> Turns out:
> 
> a) I can find only two BTS in there. There are four amplifier+duplexer wooden
> boards, but only two of them have a BTS on them. Does anyone know where the
> other two went?
> 
> b) The good news is: both BTSes that I tested are quad band, i.e. they support
> whichever band we want. Lynxis and I also each have one quad-band sysmoBTS. So
> we have at least four BTS available now, I assume it will be six in the end.
> (Details: a sysmoBTS is physically limited to specific bands. Which bands are
> supported is indicated in the EEPROM, and if you configure osmo-bts.cfg with a
> band that isn't supported, you get, during osmo-bts startup:
> "DL1C <0006> l1_if.c:1453 BTS band GSM850 not supported by hw"
> i.e. the two that worked at AfRA on sunday positively support the 850 band.)
> 
> c) To transceive with 2W, we need amps and duplexers.
> 
> The current four amps in the event pool are 1800; sysmocom has more than enough
> lo-band amp boards that we could replace in these amps with "minimal mechanical
> effort".
> 
> Sysmocom has three 850 duplexers available that we can transfer to the event
> pool.
> 
> 
> In summary:
> 
> If we get 850MHz spectrum:
> We are limited by duplexers to three sysmoBTS 850MHz on 2W, and could setup
> another one to three sysmoBTS 850MHz on 200mW.
> 
> If we get 1800MHz spectrum:
> We can have four BTS on 2W like last year, and could setup two more on 200mW.
> There are also two 1800 nanoBTS in the pool.
> 
> This is pretty good news, it means that whichever spectrum we get, we can do 2G
> with at least three 2W cells. Now let's wait/hope for at least one license
> confirmation, I think that is quite likely to happen.
> 
> Maybe we can even get both 850 and 1800, then we could do 2G on 1800 and 3G on
> 850. There are eight nano3G in the event pool, plus some more from me, lynxis,
> maybe others.
> 
> ~N



From lynxis at fe80.eu  Tue Nov 13 16:33:16 2018
From: lynxis at fe80.eu (Alexander Couzens)
Date: Tue, 13 Nov 2018 17:33:16 +0100
Subject: [osmocom-event-orga] good news everyone: BTS and bands
In-Reply-To: <20181113160428.GA18310@my.box>
References: <20181113160428.GA18310@my.box>
Message-ID: <20181113173316.0b072891@lazus.yip>

> a) I can find only two BTS in there. There are four
> amplifier+duplexer wooden boards, but only two of them have a BTS on
> them. Does anyone know where the other two went?

I'll look into the inventory, how many there should be.

> c) To transceive with 2W, we need amps and duplexers.

Last year we only used to 200mW on 2G.

> Sysmocom has three 850 duplexers available that we can transfer to
> the event pool.

Nice to hear. But do we have Antennas for 850?

> Maybe we can even get both 850 and 1800, then we could do 2G on 1800
> and 3G on 850. There are eight nano3G in the event pool, plus some
> more from me, lynxis, maybe others.

There should be 20 nano3Gs. I guess you missed the blue eurobox
(plastic).

Best,
lynxis
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181113/9fca475c/attachment.bin>

From nhofmeyr at sysmocom.de  Wed Nov 14 12:52:16 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Wed, 14 Nov 2018 13:52:16 +0100
Subject: [osmocom-event-orga] good news everyone: BTS and bands
In-Reply-To: <20181113173316.0b072891@lazus.yip>
References: <20181113160428.GA18310@my.box> <20181113173316.0b072891@lazus.yip>
Message-ID: <20181114125216.GA17000@my.box>

On Tue, Nov 13, 2018 at 05:33:16PM +0100, Alexander Couzens wrote:
> > c) To transceive with 2W, we need amps and duplexers.
> 
> Last year we only used to 200mW on 2G.

Ah, interesting.

> > Sysmocom has three 850 duplexers available that we can transfer to
> > the event pool.
> 
> Nice to hear. But do we have Antennas for 850?

Forgot to check on that. If we're fine with 200mW anyway, then we don't need to
bother...

> There should be 20 nano3Gs. I guess you missed the blue eurobox
> (plastic).

I did look, but maybe I misinterpreted the number of boxes / didn't count
enough layers / remembered the count wrongly. Most likely my count is wrong and
they are all there.

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181114/fde979cf/attachment.bin>

From nhofmeyr at sysmocom.de  Wed Nov 14 22:21:22 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Wed, 14 Nov 2018 23:21:22 +0100
Subject: [osmocom-event-orga] plan more meetings
Message-ID: <20181114222122.GD16956@my.box>

Hi GSM folks,

it's only a month until the first of us will leave for congress :)

1) what are good times for everyone to come together and sync?
please fill in: https://dudle.inf.tu-dresden.de/35c3_gsm_guru_sync/

2)
It might be nice to schedule a time where we can set up the dev core machine
together. Either remotely or in the afra. Let's see in the dudle.
- upgrade to debian testing to avoid the gcc segfault.
- polish the ansible to give us a "one-click" (TM) ready system.
- connect one or two sysmoBTS; connect one or two nano3G.
- interface with the POC / interface with a yate that simulates the POC.
- get that webservice running that the POC needs.

Peter, where is your prod machine at the moment? Is it possible to put it
somewhere where we can log in and mess with it?

3)
last sunday I didn't get through to show the interesting bits;
How interested is anyone in going into more practical details of what our GSM
setup will look like to the layman? Should we schedule another day for that?
I think it would be best to first have a dev setup running that is ideally
already interfaced with the POC, and then run a workshop on that stuff, fairly
close to congress.

...are we even reaching everyone that was there via this list?

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181114/a223f603/attachment.bin>

From lynxis at fe80.eu  Fri Nov 16 14:56:22 2018
From: lynxis at fe80.eu (Alexander Couzens)
Date: Fri, 16 Nov 2018 15:56:22 +0100
Subject: [osmocom-event-orga] good news everyone: BTS and bands
In-Reply-To: <0C51C82F-5B1A-4812-9B65-47D7A10FBAAD@ccc.de>
References: <20181113160428.GA18310@my.box>
 <0C51C82F-5B1A-4812-9B65-47D7A10FBAAD@ccc.de>
Message-ID: <20181116155622.6ea28c90@lazus.yip>

Hi Frank,

thanks for the update.

Nobody ask any other operators. I must admit, I didn't thought about it
nor do I have any contacts to them.

> I have no word from Deutsche Telekom so far. Has anyone else progress
> with other operators or the BNetzA?

We're in contact with the BNetzA to get a 850 Mhz license, but 1800
would be nice, because we have the hardware for that frequency.

Best,
lynxis
-- 
Alexander Couzens

mail: lynxis at fe80.eu
jabber: lynxis at fe80.eu
mobile: +4915123277221
gpg: 390D CF78 8BF9 AA50 4F8F  F1E2 C29E 9DA6 A0DF 8604
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181116/573c23c9/attachment.bin>

From peter at stuge.se  Fri Nov 16 15:36:50 2018
From: peter at stuge.se (Peter Stuge)
Date: Fri, 16 Nov 2018 15:36:50 +0000
Subject: [osmocom-event-orga] plan more meetings
In-Reply-To: <20181114222122.GD16956@my.box>
References: <20181114222122.GD16956@my.box>
Message-ID: <20181116153650.19409.qmail@stuge.se>

Hi!

Neels Hofmeyr wrote:
> Peter, where is your prod machine at the moment? Is it possible to put
> it somewhere where we can log in and mess with it?

It's really our (GSM team) machine, I've just taken care of it so far.

At the moment it's in the CCCV Berlin warehouse, and there's a ticket
open for transporting it to CCCB, where I would set it up to be reachable.

That ticket has been open for some time now - I'll try to bump it, but
maybe we have to resolve it ourselves. Unfortunately I'm not able to do
the transport myself, or I would have already. :\


> I think it would be best to first have a dev setup running that is ideally
> already interfaced with the POC, and then run a workshop on that stuff,
> fairly close to congress.

Agree, I think that makes good sense.


//Peter


From peter at stuge.se  Mon Nov 19 16:16:30 2018
From: peter at stuge.se (Peter Stuge)
Date: Mon, 19 Nov 2018 16:16:30 +0000
Subject: [osmocom-event-orga] plan more meetings
In-Reply-To: <20181116153650.19409.qmail@stuge.se>
References: <20181114222122.GD16956@my.box>
 <20181116153650.19409.qmail@stuge.se>
Message-ID: <20181119161630.12261.qmail@stuge.se>

Peter Stuge wrote:
> > Peter, where is your prod machine at the moment? Is it possible to put
> > it somewhere where we can log in and mess with it?
> 
> At the moment it's in the CCCV Berlin warehouse, and there's a ticket
> open for transporting it to CCCB, where I would set it up to be reachable.
> 
> That ticket has been open for some time now - I'll try to bump it

Brief update: The server is unfortunately stuck under lots of other
stuff; the warehouse crew is working on digging it out. The ticket is
not lost.


//Peter


From nhofmeyr at sysmocom.de  Tue Nov 20 00:13:09 2018
From: nhofmeyr at sysmocom.de (Neels Hofmeyr)
Date: Tue, 20 Nov 2018 01:13:09 +0100
Subject: [osmocom-event-orga] plan more meetings
In-Reply-To: <20181119161630.12261.qmail@stuge.se>
References: <20181114222122.GD16956@my.box>
 <20181116153650.19409.qmail@stuge.se>
 <20181119161630.12261.qmail@stuge.se>
Message-ID: <20181120001309.GC18764@my.box>

On Mon, Nov 19, 2018 at 04:16:30PM +0000, Peter Stuge wrote:
> Brief update: The server is unfortunately stuck under lots of other
> stuff; the warehouse crew is working on digging it out. The ticket is
> not lost.

Brief update from the ansible side: I am working on the 34c3 ansible setup on a
neels/wip branch in lynxis' git at code.fe80.eu:congressgsm/34c3.git

Changing it to use osmo-dev's top-level makefile and config templates.

Also adjusting to use the systemd *.service files that are now part of the
osmocom git repositories, installed at 'make install' time. Since they assume
binaries in /usr/bin, I am also installing with --prefix=/usr now, not under
/srv/34c3/. For the same reason I want to place the config files in
/etc/osmocom/, possibly by a symlink to /srv/34c3.

One gotcha is, osmo-mgw installs osmo-mgw.service, and I think we'd rather like
to have osmo-mgw-for-bsc.service and osmo-mgw-for-msc.service, i.e. two of
them. So I think I'll add to the ansible some cp and sed to make two separate
services from the installed osmo-mgw.service file. I mean, it would work fine
with just one osmo-mgw for both BSC and MSC, but if we want to untangle an RTP
stream from the logging, having a separate instance helps tremendously.

Ah, yeah, and we might want to rename 34c3 in there to 35c3, or maybe rather
call it neutral 'event' or 'gsm' or 'c3', so we don't have to rename for every
event.

The ansible so far works pretty nicely in an lxc.

But so far all is root, installed as root, run as root. My goal is to have a
'gsm' user instead to run the components, out of basic sanity.

Still waiting for that VPN to the POC machine. So far I only got a key and
don't know how to use it :P

~N
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.osmocom.org/mailman/private/osmocom-event-orga/attachments/20181120/1e8385b8/attachment.bin>