10:44:25	 From  Michail Dalgitsis (NBC) : Is Open RAN studio open-source?
10:45:22	 From  Alireza MOHAMMADI (BubbleRAN) : Hi @Micahil -- Even though Open RAN Studio enables use of Open Source software like OAI or SRS, itself is not Open Source.
10:46:08	 From  Michail Dalgitsis (NBC) : Reacted to "Hi @Micahil -- Even ..." with 👍
10:46:41	 From  Fatemerajabi : Replying to "Hi @Micahil -- Even ..." 

 But in your blog you...
10:47:15	 From  Anup, TU Delft : Are there specific packages of Open RAN Studio for Education/ Research in Universities?
10:47:28	 From  Alireza MOHAMMADI (BubbleRAN) : The code is not open for access, but for usage it is via a subscription service that comes with considerable discounts for education.
10:47:51	 From  Alireza MOHAMMADI (BubbleRAN) : Look for the difference between Open Source and Free Software
10:48:36	 From  Mayank : Replying to "Are there specific..."

+1
10:50:04	 From  Fatemerajabi : Replying to "Hi @Micahil -- Even ..." 

 I got what do you me...
10:51:04	 From  Navid Nikaein : Dear All, The recorded video will be published on the BubbleRAN Youtube channel. We will send a separate email with all the materials.
10:51:21	 From  NTNU : Reacted to "Dear All, The record..." with 👍
10:51:23	 From  Fatemerajabi : Reacted to Dear All, The record... with "👍"
10:51:31	 From  Amitava Middya : Reacted to "Dear All, The reco..." with 👏
10:51:35	 From  Amitava Middya : Removed a 👏 from "Dear All, The reco..."
10:51:37	 From  Amitava Middya : Reacted to "Dear All, The reco..." with 👍
10:51:39	 From  Tori : Reacted to "Dear All, The record..." with 👍
10:51:41	 From  Tori : Removed a 👍 reaction from "Dear All, The record..."
10:51:44	 From  Amitava Middya : Removed a 👍 from "Dear All, The reco..."
10:51:45	 From  Tori : Reacted to "Dear All, The record..." with 👍
10:51:45	 From  Amitava Middya : Reacted to "Dear All, The reco..." with 👍
10:55:09	 From  Anderson Bravalheri - University of Bristol : I suppose Free5GC is also supported, right? (Given that there is a talk about free5GC later today...)
10:56:18	 From  Navid Nikaein : for this release we are not supporting it, but we if we see needs from the community we will on-board Free5GC too
10:56:38	 From  mojtaba : Reacted to "Dear All, The record..." with 👍
10:56:43	 From  Anderson Bravalheri - University of Bristol : Reacted to "for this release we ..." with 👍
10:56:48	 From  tv20908 : Reacted to "for this release we ..." with 👍
10:57:05	 From  Navid Nikaein : This is said, we support Amarisoft RAN/CN/IMS, not in the MX-ORS, but on the other products
11:00:12	 From  Alireza MOHAMMADI (BubbleRAN) : The product was free for the early adaptors only and now we offer discounts for education.
11:00:41	 From  Alireza MOHAMMADI (BubbleRAN) : Yes. Please be in touch with us through email or website and we send you the details.
11:01:16	 From  Manimaran Chandran : Replying to "Hi @Micahil -- Even ..."

If we take it for education purpose there will be a dedicated support to build up whole eco system?
11:01:45	 From  Bharat Agarwal, Vodafone Tech Innovation Center : is this R1 interface is new?
11:01:52	 From  Navid Nikaein : Replying to "Hi @Micahil -- Even ..."

It is the same model as RedHat and Ubuntu that package and release Linux with added value features and enterprise-grade support level.
11:02:22	 From  Amitava Middya : Reacted to "for this release w..." with 👍
11:03:06	 From  Navid Nikaein : Dynamic xApp are those xApps that are dynamically deployed by OAM (here Athena) based on the policy that is enforced by the non-RT RIC (here ODIN)
11:03:23	 From  Alireza MOHAMMADI (BubbleRAN) : R1 standardization is ongoing at O-RAN. This is BubbleRAN's take and approach to it, close to the specifications made by O-RAN.
11:03:48	 From  Anderson Bravalheri - University of Bristol : So my understanding is that ODIN works like a Kubernetes operator
11:04:14	 From  Navid Nikaein : Both Athena and Odin ate K8s Operators
11:04:21	 From  Anderson Bravalheri - University of Bristol : Reacted to "Both Athena and Odin..." with 👏
11:05:10	 From  Michail Dalgitsis (NBC) : is the Dynamix App deployment defined on the O-RAN specifications or it is a feature/capability your framework supports?
11:05:39	 From  Michail Dalgitsis (NBC) : Replying to "is the Dynamix App d..."

Dynamic xAPP*
11:07:06	 From  Alireza MOHAMMADI (BubbleRAN) : ODIN is implemented as a Kubernetes Operator but acts as Non-RT RIC in O-RAN stack
11:07:51	 From  Alireza MOHAMMADI (BubbleRAN) : It is a modern terminology that we have adapted. We see it in more detail in the second half of the episode.
11:09:15	 From  Anderson Bravalheri - University of Bristol : I understand... using kubernetes as a framework for the development of the controllers...

I am also assuming xApps are somehow represented as CRDs right?
11:12:02	 From  Michail Dalgitsis (NBC) : can we zoom in the screens?
11:13:08	 From  Alireza MOHAMMADI (BubbleRAN) : The descripton of the xApps is defined as CRDs. The xApps themselves are containers.
11:13:42	 From  Alireza MOHAMMADI (BubbleRAN) : Please use the Zoom features to increase the screen size if required.
11:14:44	 From  Michail Dalgitsis (NBC) : Reacted to "Please use the Zoom ..." with 👍
11:16:09	 From  shadi Moazzeni : What will be the role of Athena here?
11:16:55	 From  Anderson Bravalheri - University of Bristol : Reacted to "The descripton of th..." with 👍
11:17:03	 From  Alireza MOHAMMADI (BubbleRAN) : OAM (Operation and Maintenance) according to O-RAN terminology.
11:17:16	 From  Alireza MOHAMMADI (BubbleRAN) : MANO in terms of ETSI.
11:17:17	 From  Anderson Bravalheri - University of Bristol : Replying to "I understand... usin..."

I understand, we use a similar approach in our lab.
11:17:39	 From  tv20908 : Reacted to "The descripton of th..." with 👍
11:17:44	 From  Andrea Marotta (UNIVAQ) : How can we define slice requirements? Bandwidth only?
11:17:49	 From  Navid Nikaein : Interactive xApps are those that a developer can interact directly using a terminal. They expose a set of high level commands to interact with the RIC/E2 Agent infrastructure  (multiple of those)
11:18:43	 From  shadi Moazzeni : Reacted to "OAM (Operation and M..." with 👍
11:19:20	 From  Majid Salim : If we have multiple Networks, how can we choose the right network for connecting UEs to it?
11:19:27	 From  Navid Nikaein : Reacted to "I understand, we use..." with 👍
11:20:15	 From  aykut : Reacted to "How can we define ..." with 👍
11:20:39	 From  Alireza MOHAMMADI (BubbleRAN) : There is a typo in the image:
targetPrbUsg is the correct key. He would use the right name though in the file.
11:21:05	 From  Alireza MOHAMMADI (BubbleRAN) : Slicing demo is on the second part. Please stay tuned!
11:21:53	 From  Alireza MOHAMMADI (BubbleRAN) : You choose the core networks and the preferred access name in the YAML input by the names that you have already defined in the other steps.
11:22:03	 From  Constantinos Vrontos : Can the RIC directly expose and calculate metrics such as % packet drop and number of retransmissions? Or raw metrics can only be obtained and then post-processing is needed to calculate them?
11:22:46	 From  Nikos Bartzoudis (CTTC) : What other type of metrics can be extracted by the MAC, RLC and PDCP, etc?
11:23:38	 From  Nikos Bartzoudis (CTTC) : Have you tested the same example we see now with a bigger number of UEs?
11:23:48	 From  aykut : You don't have  PCF which manages all policy in the core network. How will you synchronize O-RAN policies with core network side policies?
11:25:11	 From  Andrea Marotta (UNIVAQ) : Replying to "How can we define sl..."

Thanks!
11:25:49	 From  Chieh-Chun/Eurecom : Replying to "Can the RIC direct..."

the calculated metrics can be exposed by RIC via KPM SM, if the RAN functions are implmented to calculate the metrics, ex: throughput, used RB number. In this static xApp (graphic one) we are showing here is using MAC SM (provided by FlexRIC) and do the required calculation in the xApp.
11:26:43	 From  Shah Zeb : Replying to "Have you tested the ..."

I have also same question. What about the large number of UEs especially the PRBs utilization in maximum users scenario?
11:26:48	 From  Bharat Agarwal, Vodafone Tech Innovation Center : When UE connected to slice 0 why slice 1 PRB utilization increases?
11:28:05	 From  Navid Nikaein : Replying to "Can the RIC directly..."

So the answer is yes as long as the RAN function implements those metrics.

Alternatively, another strategy is to get all the raw stats and then calculates metrics at the xApp level.
11:29:02	 From  Chieh-Chun/Eurecom : Replying to "What other type of..."

These metrcis are all supported in OAI RAN by using SMs provided from FlexRIC, including customzied SMs (MAC, RLC, PDCP) and O-RAN SM (KPM).
11:30:31	 From  Theodoros Tsourdinis (SU/UTH) : For the PRB allocation did you use the SLICE SM? If yes, what kind of slices did you enforce? ENVS?
11:31:15	 From  Constantinos Vrontos : Reacted to "So the answer is yes..." with 👍
11:31:17	 From  Constantinos Vrontos : Reacted to "the calculated metri..." with 👍
11:31:39	 From  Navid Nikaein : Replying to "What other type of m..."

As mentioned by Chiehchun, we have specified custom service model for each layer of RAN, from which you can get all the detail statistic of PDCP, RLC, MAC, GTP, and RRC
11:32:06	 From  Navid Nikaein : Short 5 minutes break
11:34:02	 From  Alireza MOHAMMADI (BubbleRAN) : There are differences between slice at the level of core network at slice at the level of the RAN. RAN is simply slicing the "radio resources". Of course there needs to be coordination and that is what ODIN tries to achieve by contacting both RIC and the CN.
11:35:11	 From  Ilias Chatzistefanidis (EURECOM) : The PRB utilization is scoped on the UE for this demo. The 2 slices have different capacities (slice0: 50%, slice idle: 5%). So when we associate UE to slice idle, we see the UE PRB reduce and when we come back to slice0 we see the UE PRB peak
11:36:13	 From  Yi Chu (University of York) : Does FlexRIC support RC? I saw your RIC logs have RC SM loaded but RC is not mentioned on FlexRIC docs
11:36:15	 From  Ilias Chatzistefanidis (EURECOM) : yes SLICE SM and we used NVS with CAPACITY
11:38:52	 From  Navid Nikaein : Replying to "Have you tested the ..."

In MX-ORX, we support up to 4 soft UEs per gNBs for stability purpose in OAI (different value for srsRAN). This could go much higher as long as you have enough compute power. 

Note that there is difference between connected and active UEs. This means that even with 4 active UEs you could reach maximum PRB utilization.
11:39:04	 From  Michail Dalgitsis (NBC) : Replying to "For the PRB allocati..."

what is SLICE SM? it is a standardized xApp?
11:41:12	 From  Theodoros Tsourdinis (SU/UTH) : Reacted to "yes SLICE SM and we ..." with ❤️
11:43:00	 From  Ilias Chatzistefanidis (EURECOM) : In other words, we aggregate the PRB of both slices and calculate the sum
11:44:17	 From  Chieh-Chun/Eurecom : what is SLICE SM? it is a standardized xApp?
it's a customized service model defined by FlexRIC to control the slices in the level of scheduling radio resource.
11:44:29	 From  Ilias Chatzistefanidis (EURECOM) : SLICE SM is our custom service model (SM) for applying slicing control on the RAN from the xApp
11:47:41	 From  Hanif Kukkalli : Are all these pods running on single k8s cluster or are they running on multi cloud environment?
11:48:37	 From  Navid Nikaein : Replying to "Are all these pods r..."

They are running on a single cluster composed of 11 compute nodes on a private cloud.
11:49:18	 From  Hanif Kukkalli : Reacted to "They are running on ..." with 👍
11:52:32	 From  Alireza MOHAMMADI (BubbleRAN) : @Nikos: OAI currently supports up to 4 UEs on the same gNB. Utilizing some tricks you could make it more in cost of losing the level of realization. The current demo as you see shows a couple of UEs per gNB.
11:52:46	 From  Hanif Kukkalli : Replying to "Are all these pods r..."

Apart from research purpose, using k8s, is that really scalable and usable when we want to deploy lets say just one gNB at one edge site connected to disaggregated core network like having multiple AMF/SMF/UPFs across a geographical area?
11:55:05	 From  Hanif Kukkalli : Replying to "Are all these pods r..."

k8s also has port exposure range which forces us to rewire exposed ports of each different service in each k8s cluster.
Is there a way to mitigate this?
11:55:55	 From  Alireza MOHAMMADI (BubbleRAN) : K8s is used in production for several projects and products already.
11:56:27	 From  Alireza MOHAMMADI (BubbleRAN) : There are different distributions of K8s enabling different features. To mention a few: Openshift, Rancher, GKE
11:57:04	 From  Alireza MOHAMMADI (BubbleRAN) : Building on top of any of those distributions, our SMO addresses all the challenges for 5G/6G/Open RAN, including the ones you mentioned.
11:58:02	 From  NTNU : E42 and R1 interfaces are specified by O-RAN Alliance?
11:58:36	 From  Alireza MOHAMMADI (BubbleRAN) : R1 is an interface placeholder in O-RAN, e.g., it is not fully defined yet but it is specified.
11:58:58	 From  Alireza MOHAMMADI (BubbleRAN) : E42 is the mirror of E2 APIs for the xApps and it is particular to FlexRIC.
11:59:49	 From  Behnam Ojaghi : Very fruitful demo and discussion, many thanks.
11:59:59	 From  NTNU : Replying to "E42 and R1 interface..."

Thanks :)
12:00:35	 From  Navid Nikaein : Reacted to "Very fruitful demo a..." with 👍
12:01:02	 From  Michail Dalgitsis (NBC) : Reacted to "SLICE SM is our cust..." with 👍
12:02:09	 From  Michail Dalgitsis (NBC) : Replying to "For the PRB allocati..."

the custom SMs for xApss and the FlexRIC are all opensource and come with documentation and tutorials to run them on Kubernetes?
12:02:16	 From  Hanif Kukkalli : Reacted to "Building on top of a..." with 👍
12:02:21	 From  Hanif Kukkalli : Reacted to "There are different ..." with 👍
12:02:24	 From  Hanif Kukkalli : Removed a 👍 reaction from "There are different ..."
12:06:27	 From  Navid Nikaein : Replying to "For the PRB allocati..."

Partially yes, please visit here:

https://gitlab.eurecom.fr/mosaic5g/flexric
12:06:37	 From  Michail Dalgitsis (NBC) : Replying to "For the PRB allocati..."

thanks
12:12:33	 From  Chieh-Chun/Eurecom : Replying to "E42 and R1 interfa..."

in the O-RAN terminology, the E42 interface in here provides the same functionalities as E2-related APIs from NearRT-RIC APIs.
12:28:21	 From  Nikos Bartzoudis (CTTC) : How about using eBPF for secure and fast data collection? Do you contemplate such type of extension?
12:28:31	 From  VG, IITDh : Great sessions!
Can we get the recordings of these session?
12:28:44	 From  VG, IITDh : *sessions?
12:28:49	 From  Mikel (OAI) : Did you see any improvement when you assign more threads in the Task Manager of the nearRT-RIC?
12:30:03	 From  Alireza MOHAMMADI (BubbleRAN) : @Nikos: We're using eBPF for the energy metric collection. Energy metrics could be calculated generally and external to the gNB. Accessing the acutal metrics in the RAN is a more sophisticated task that needs E2.
12:30:50	 From  Alireza MOHAMMADI (BubbleRAN) : Recordings would be uploaded on YouTube of BubbleRAN
12:31:00	 From  Navid Nikaein : Replying to "Great sessions!
Can ..."

It will be on our your tube channel. 

We will share the link next week.
12:31:06	 From  Navid Nikaein : Reacted to "Great sessions!
Can ..." with 👍
12:31:47	 From  Michail Dalgitsis (NBC) : has BubbleRAN any open-source version?
12:32:50	 From  Michail Dalgitsis (NBC) : also, is BubbleRAN disaggregated gNB different than the DU/CU of OAI?
12:33:10	 From  Navid Nikaein : Replying to "has BubbleRAN any op..."

No. We do have a pay as you go model for Open RAN Studio.

Please have a look here: 
https://bubbleran.com/products/mx-ors/
12:33:11	 From  Alireza MOHAMMADI (BubbleRAN) : In this demo we used the default (2). We try later to use more and I need to set it also at the OAM to reserve the CPUs.
12:34:00	 From  Alireza MOHAMMADI (BubbleRAN) : I'll ask Khoa for his demo how many he used
12:34:42	 From  Alireza MOHAMMADI (BubbleRAN) : I'll ask Khoa how much he used for his demo on bare-metal
12:35:20	 From  Mikel (OAI) : Thanks!
12:35:49	 From  Alireza MOHAMMADI (BubbleRAN) : It is based on it but with minor improvements, including better support for the UEs and E2, support fo slicing, and more stability.
12:36:26	 From  Yi Chu (University of York) : Does FlexRIC support RC? I saw your RIC logs have RC SM loaded but RC is not mentioned on FlexRIC docs
12:36:45	 From  Alireza MOHAMMADI (BubbleRAN) : Apparently, Khoa was using 2 as well.
12:38:30	 From  Chieh-Chun/Eurecom : Replying to "Does FlexRIC suppo..."

Yes, it's supported RCv1.
12:39:29	 From  mutasem hamdan (Samsung  R&D UK) : Thanks for orgnizing this even . Hasve a nice day!
12:40:12	 From  Nikos Bartzoudis (CTTC) : Thanks a lot indeed for the comprehensive session today
12:40:37	 From  Selcuk (Samsung Research UK) : Thank you for the great presentation.
12:40:43	 From  Nikos Bartzoudis (CTTC) : Kudos to all the team
12:41:25	 From  Raja Sana Gul : Thank You for this event.
12:41:33	 From  Swarna Chetty(YO-RAN, University of York) : Brilliant session, Thanks a lot.
12:41:34	 From  Shah Zeb (Uni. of Bristol, UK) : Thanks to the team for the interactive tutorial presentation on ORAN studio.
12:41:58	 From  Navid Nikaein : Reacted to "Thanks to the team f..." with 👍
12:42:01	 From  Navid Nikaein : Reacted to "Brilliant session, T..." with 👍
12:42:35	 From  Sibel Malkos : Thank you for organizing this event and giving this opportunity today. Thank you all team.
12:42:40	 From  Navid Nikaein : Reacted to "Thank you for organi..." with 👍
12:42:44	 From  Navid Nikaein : Reacted to "Thank You for this e..." with 👍
12:42:47	 From  Navid Nikaein : Reacted to "Kudos to all the tea..." with 👍
12:42:50	 From  Navid Nikaein : Reacted to "Thank you for the gr..." with 👍
12:42:54	 From  Navid Nikaein : Reacted to "Thanks for orgnizing..." with 👍
12:49:30	 From  mojtaba : Thank you for this event