Hyperledger Foundation Meetups, 23 Feb 2023

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: IT Service Management (ITSM) for Hyperledger Fabric

Description

A virtual meetup hosted by Hyperledger Budapest on February 23, 2023 featuring Daniel Szego speaking about IT Service Management (ITSM) for Hyperledger Fabric.

In this meetup we will investigate the possibilities and challenges of integrating Hyperledger Fabric with professional IT Service Management methodologies, like with ITIL (Information Technology Infrastructure Library). We stay close to the technology though and focus mostly on the Fabric specialities at different service management practices, like:

- Capacity and Performance Management
- Change Control
- Problem Management
- Incident Management
- Servie Desk
- Sevice Configuration Management
- Service Design
- Deployment manegement
- Infrastructure and platform management

The meetup can be regarded as a starting point as well to get more community involvement in this highly relevant topic.

----
Hyperledger – Open Source Blockchain Technologies

Hyperledger is an open source community focused on developing a suite of stable frameworks, tools and libraries for enterprise-grade blockchain deployments.
It serves as a neutral home for various distributed ledger frameworks including Hyperledger Besu, Fabric, Sawtooth, Indy, as well as tools like Hyperledger Avalon, Cactus and libraries like Hyperledger Aries, Ursa. Learn more about Hyperledger projects: https://www.hyperledger.org/use

Discord: https://discord.gg/rVCn4j7UFd
Case Studies: https://www.hyperledger.org/learn/case-studies
Training & Certification: https://www.hyperledger.org/learn/training
Tutorials: https://www.hyperledger.org/use/tutorials
Webinars: https://www.hyperledger.org/learn/webinars
Events: https://www.hyperledger.org/events
Vendor Directory: https://www.hyperledger.org/use/vendor-directory

Subscribe to the Hyperledger Newsletter: https://www.hyperledger.org/newsletter

Follow-us on Twitter @Hyperledger

Learn about Hyperledger Membership: https://www.hyperledger.org/about/join

#Hyperledger #Blockchain
License
Creative Commons Attribution license (reuse allowed)

A

A

Yeah you're all set Daniel feel free to get started.

A

So awesome uh so hi everybody. uh Let me just slowly start. uh This is Daniel, and this is hyperledger Meetup. uh We're gonna talk today about like Elizabethton IIT service management, itsm for hyperledger fabric or for for Consortium blockchain at all.

A

ah This is gonna, be like more like a kind of a presentation at the beginning of from me, and after that we can have like Community discussion based on that uh I won't share, like you know, best practice, experience or probably not even like you know, ready to ready to go Solutions so so might, as you might as well.

A

It's just actually an initial uh consideration or an initial uh presentation how hyperledger fabric should be considered, uh especially if you just you know, I mean consider professionality service management, somehow uh with with Consortium blockchains um at the end. uh So first, uh let me just sorry a little bit. uh My network is not very stable, usually uh I try to slow down, sometimes, especially if I feel like that. My network is, is getting a little bit buggy and it might as well happen that.

A

For 20 seconds, usually it won't stay longer, 30 seconds, so I come back and then I am pretty I got the routine a little bit so so I can continue where I was uh but again. So today the topic is: uh supposing we got high pressure fabric somewhere and we want to integrate somehow with professionality service management, which are the topics uh that should be considered uh and again um I'm not gonna give sometimes you know ready to go. Solutions rather just raise questions.

A

So that's gonna, be the agenda for today and I will start with the level and art of decentralization as usual.

A

It might be boring, but I I mean I, mean I would say it's just important, always to have somehow the roadmap uh always always to have the the point of view.

A

uh What we are talking about- it's just just much easier to to position ourselves, basically um in about uh with that, so even if it might be boring, I just start with level of decentralization some overview of decentralization and then just a couple of slides on ID service management or ITI, which is which is one way of doing like professional information, uh service management, professionality storage, service manager, then I get one slide on on Consortium blockchain applications and then I will have a couple of slides like uh on General management practices and service management.

A

Practices of ITIL, uh which are you know, I mean more like technical oriented uh and not so much holistic than other ITIL approaches, I'm more like a technical guy. Basically, so um during my presentation, I just try to stay close basically on on technology uh and not go very much into the direction of the holistic point of view of clients, for instance.

A

So let me start the usual picture. The usual uh slide. It's centralized decentralized, uh distributed and stuff like that and basically as usual, so we can consider something which is a level of decentralization so like a system can be centralized or fully centralized, something which is fully decentralized, fully trustless and basically a type of lager. Most of the Technologies are somewhere somewhere in between. So usually they are like semi-decentralized or semi-trusted systems.

A

But perhaps one important point of view is that uh so usually there are like two Dimensions. um We can usually, if we speak about like decentralization, we think of the business Logic the application itself. So that's the that's the first idea. If we get like a fully decentralized business application uh or or is it like semi-decentralized or is it like fully fully centralized? But there are other aspects as well that we can consider, and so, for instance, one less well-known. But despite the important topic is governance.

A

Means something if you get like a decent life Network. How do we change the app uh and some some parts of this app or some parameters of this app? Can change, then how we can change it practically?

A

uh So it's it's important to point out that uh if we speak about like decentralization uh business, logic or application itself can be decentralized or semi, semi-decentralized or centralized as well, but totally independently from that. The governance of that app of that Network can be again centralized decentralized or semi-decentralized or or semi-summitrustive. Basically, and this can be like two different dimensions- um not necessarily totally independent, but I mean theoretically, they might as well be considered as totally independent uh Dimensions as well, which is usually not regarded very much in in live blockchain system.

A

That's that's Operation. That system operation and I would like to point out that, basically, governance is not necessarily the same as operation, so like operating a system has like a a lot of like very practical things. So one practical thing is like analyzing the log, so governance is usually something which is you're, not changing the parameters. It's it's sometimes much more on the on the on the on the on the theoretical basis uh for operation. We got a lot of practical stuffs as well, so I would say.

A

Governance is like subset of operation because, of course, I mean for operation. uh I would consider. Basically, the changing parts of the system is part of governance as well. But if we speak about like operation or system operation, there are many very, very practical and pragmatic stuffs and things that should be carried out again.

A

Analyzing log uh taking look on monitoring and deciding if the system is okay or if it's running or not running and and other things as well, uh so that's Operation, uh and then it's important to point out that, basically, if we speak about like distance.

A

Especially like this is Operation, uh we can have many variation, even if that is this decentralized. So, even if you like have like a governance in a decentralized way, uh we can do it totally independently from blockchain from or from any IIT system. So, just like you know, I mean consider. Basically it's like a business network of of like 10 universities uh and then, if they want to decide something in a totally decentralized way, then they might just set up. You know call for instance, so it can be something which is totally of jail.

A

uh We can consider like of chain governance, or perhaps even of ID uh or or offline uh governance as well and operation as well, and we can consider something which is like on chain, governance or or own chain supporter support governance as well. So if you speak, I speak about like decentralization, especially of governance and operations,.

A

Operation and we can have something as automated or manual governance as well so like. If you got a change in a system and practically it can be carried out in a menu away, uh it can be carried out or executed in a more automated way, and if your governance model is fully decentralized, it can even carried out in a in an automated and decentralized way as well.

A

So these are all possibilities and again so we usually speak about like the decentralization of business logic, but especially if we want to operate, maintain such a business Network or such a decentralized network. What's even more important is the art of decentralization or the level of decentralization of governance and operation and the support uh the automation, uh which is uh which helps basically uh for this for this operation, so I decide basically to collect a couple of system uh based on this two Dimensions. So we got here on this picture.

A

We got two dimensions: on the one hand, we got the application Level, which is pretty much the business logic, so we got like centralized application on one end and decentralized applications on the.

A

Let me say: business networks or business blockchain Network, which is kind of a semi-trusted semi-decentralized way of of of of of of ID and on the other dimension, I got something which is which is the governance or the operation again. I mean governance and operation are not exactly the same, but I mean for sake of easiest I. Just I just have one dimension, so we got here something which is a decentralized governance or decentralized operation and on the other hand we have something which is a centralized governance and centralized operation.

A

So I got to several a couple of examples here, and so usually you might as well see uh hey if you get. If you got like a decentralized business logic, then we need like decentralized governance or decentralized operation as well, but this is not a must, so this usually happens that way, but it shouldn't work always that way, uh especially not with consolation blockchains.

A

So I got several examples here. We got the public blockchains, for instance, here so like Bitcoin, is very, very decentralized as an application, uh but if we just consider the governance and operation, it is very decentralized, but it's difficult it's difficult to get get the consensus or concerns from the people uh running notes. For instance, there's always a very, very tough discussion if, like a hard, Fork should be carried out and so and so on and, for instance, in Bitcoin. This whole decentralized governance and operation is not supported by the Bitcoin itself.

A

It should be done by I, don't know different, chats or or or or or or other uh um with other ideas, instruments or other stuffs like that is certainly not so decentralized as Bitcoin, but, for instance, in terms of governance and operations, uh tesos practically got something which is an own chain governance, so it supports certain activities. Certain changes of the network with like uh on chain voting and basically on chain kind of a change mechanism practically so.

A

For this reason, I just put basically in terms of governance or operation in a little bit like more decentralized than Bitcoin, is a little bit less decentralized as Bitcoin. Basically, and it doesn't have something which is like uh you know: I mean on chain governance and operation and I just have one more example: that's EOS EOS, if I'm not mistaken here since, like 21 notes, which.

A

So basically I would say it's not as much uh decentralized the CT, remote autos and so on, and we got somebody here. The the consumption, blockchains um and, of course, I mean. If you speak about like conversion blockchains, then they are like semi-trusted networks so somewhere in between and usually the the main idea is that say, I mean we got. We got some interested business applications and then we got like semi-trusted uh governance or operation model, but this is not necessarily the case.

A

I mean if you consider the classical examples like hyperledger, Fabric or another one is like blockchain and Identity or our free Corner. uh These are actually platforms uh which you can many different platforms, many different scenarios, so you can have like platforms even with Hyper, larger Fabric in a more decentralized, more semi-trusted level, but you can basically uh go and and centralize several components you can have. You can have like keyboard networks.

A

You know in a way that part of your network is like semi-decentralized, and there are some points which are fully centralized and so on, and so so I got to a lot of bubbles uh because you know I mean usually, if you, if you have a Consortium blockchain uh the more requirement that you would need like something which is at least a semi-trusted network for your business application.

A

But it is not necessarily the case that you need the governance or operation as well in a in a fully I mean not fully, but at least a semi-decentralized way. It might happen, for instance, that in a business corporation uh there are.

A

Decentralized, but the the operational governance shouldn't be actually such uh such so decentralized, so it makes things, makes things easier, I'm, just just giving one example, which is which is not a blockchain example. It's like the uh the the chips that work, uh if I'm not mistaken, uh among Among Us Banks, basically,.

A

So I think Daniel's just having some audio or network issues, I think he had said that he'll be back in a minute or two if he does drop. So let's just give him a minute yeah that was my 22nd break.

A

Actually, so what I wanted to point out that basically I mean, even if you say, hey, blockchain and that's it's decentralized, I speak about like business cooperations and then not necessarily. All aspects should be equally decentralized, specifically like like governance and operation. uh It might be just simpler to to to getting like more more in a centralized world. So that's the overview of decentralization and again so practically.

A

If you speak about like professionality service management for a Consortium blockchain, it is always important to decide if we, if we want to realize as decentralized as possible or or if, if we say for certain items. Third, certain elements: hey we don't we don't necessarily I mean we need the decentralization for the business application for the business logic itself, but we don't necessarily need the fully decentralized uh way of thinking for, for the, for the managed for the ID search for the operation set.

A

uh Just going back like if you did I mean uh two more two more two more actually aspects, so things getting more complicated. Of course, if we start to speak about like like Hebrews blockchain Solutions, uh then it's going to be even more complicated. uh That's for sure. um So we will see how the word where it's going to develop. If we, if we had like uh hybrid blockchain, Solutions.

A

Practically uh to to customers- and there are like many methodologies for for IIT service management- uh there's an evolution as well in in like perhaps a competition as well in different different IIT service management, Technologies.

A

um So one one way, it's practically probably a little bit more classical that's ITIL or I2 Information Technology infrastructure library. But there are other methodologies as well like the business process, framework or control objective or the copies that's for for telecom companies, probably and then especially lately, we got a lot of like you know: I mean which are partly it service management methodologies. It's like doing everything, agile, doing more, like devops or or secure devops devsecops, if I'm, not mistaken or lean, and so on, and so on, and so on.

A

uh So I just here, I feel uh which is yeah again I'm, not graduates, it's like more classical uh but.

A

uh It's it's much better to think of often uh of of a little bit like classical uh I.T service management, technology uh or methodology like ideal, and it's it's more difficult to integrate everything like into devops. So devops work pretty well. If you get one company, a couple of front-end databases and so and so on, but having like a multi-organization setup, uh then my experience that these These Old Fashions methodologies suit a little bit better. But again, that's just my experience.

A

So if you think like hey, uh high blood, Fabric and devops and agile is the best way of doing things that you know, I mean just just share the experience. So uh don't worry so so idea: Information Technology infrastructure, Library. uh That has many versions. That's the it started like in 2000, if I'm not mistaken, so we got like the version, one two and three and four and so version three was pretty compact um version. Free is a little bit like more holistic and value-centric uh methodology.

A

Which are I mean very important. It's more like you know, on a holistic and philosophical level. It is a lot of things like value value Creations, where you've seen how how organizations should behave, how people should actually collaborate and be educated in uh in the in the in the in the organization uh and a lot of stuffs like like costing risk and utility of a service and so on and so on, which are very much important again.

A

They are not not so technical for me, so what I'm gonna focus on basically in the next couple of slides. These are the management, practices and I think these management practices can be found actually in any other uh methodologies as well, so I hope it's it's it. They gonna like act as a good I mean core understanding. uh If, if we speak about like like operation or uh professional professionality management and hypoallergic fabric, um one one thing to consider.

A

That's usually overseen, so we speak about like I mean I, mean operating or maintaining, or doing like professionality uh service management for high collagen Fabric or for a Consortium blockchain, but usually uh I mean that's, not the scope, uh because I mean like. Let me just have hyper Ledger fabric hyperledger fabric usually doesn't stay like alone in the company uh you I mean fabric won't be enough uh alone uh for you to build up a complex application for a complex application.

A

There are a couple of other things that should be considered and then I just have like two other things here. So uh first I mean we got the infrastructure because I mean fabric should run somewhere and then, if you speak of like infrastructure, we got network storage, virtual virtualization.

A

uh It's actually like, uh like a like a uh very more than me,.

A

So probably, you need something as as orchestration uh below hyperledger, Fabric and I, just put like the couples to Virtual Machine, what kind of a microservice architecture, uh but nevertheless I mean this infrastructure part is pretty complex and simply it is required that we that we have hyperledger fabric uh up and running and for a full scale application. We need usually other services as well. So let me just imagine kind of an application which has user users I mean uh just directly communicating somehow with the with DVD network.

A

uh So in a simplest case we get kind of a user experience which might have, like you know, I mean components as as UI front-end, uh perhaps centralized business logic or or centralized business logic, Services, perhaps centralized apis and other way of of communicating steps.

A

So what I want to point out, basically, uh is that usually, if you speak up of like IIT service management and typological fabric, uh we should consider that that's, that's, probably a full application uh that should be up and running not just fabric.

A

uh So it's much better to speak of you know: IIT service management for an application that has fiber hyperledger fabric uh somewhere in between as well, which is, of course, probably the most difficult part. But despite uh it shouldn't be forgotten that we got some some other components as well. uh Below on the infrastructure, side and and upper I would say upper on the on the user side as lab.

A

So let me just go on. Let me just consider some some some management practices uh and again in the next, like six slides, I'm, just gonna list, a couple of light management practices and some some considerations so from the general management practices. One one thing is is usually the continual uh Improvement.

A

From the theoretical side, it is usually not so easy to realize from the Practical point of view, I mean that's my experience, so this usually means if you get like an application that should be practically somehow uh improved. uh And then, if you get like very devops methodologies, usually devops say that hey I mean I mean we can even uh change. We can even deploy use something uh practically on a daily basis. That's that's not such a problem.

A

So a continual Improvement is something which is you know, I mean if we speak of an application site, it's something which is develop and deploy so I would say changing the business logic uh or something which is more infrastructure Centric. It's like plan rollouts and operate uh for instance. So if you speak of like.

A

Is that uh we should we should somehow split get a component like like like the chain code itself, I mean that's the smart contract for for hyperledger fabric that behaves more like an application like a business application. You know it contains the business logic and the business logic. So I must say, I mean I, mean I, mean business. Lodging shouldn't actually change very often in blockchain.

A

uh So it's it's, certainly not a good idea to to release like even even Jinx, with our smart contract on a daily basis, but in a Consortium blockchain it might change more frequently. Basically, so the chain code part is something which is more similar to an application. Basically and I would say all my experience is that the other parts of fabric which is like you know the piece, the orderer, the certificates authorities, the coach, Davies and stuff like that they behave more like an infrastructure.

A

So even if you consider an infrastructure that should be actually somehow uh refreshed or continually improved uh can be a good idea to keep like hyperledger fabric on the on the latest release. So you got like uh like in every few months. uh Basically, a new release, a new minor release. uh So, if you you know, I mean want to do it in a very professional way.

A

Then, even if, like that's that's a kind of fabric infrastructure, it's a good idea to uh to just um you know: I mean I, mean release or deploy the new fabric release, which is again around three months uh and I must say so. Yeah sometimes.

A

Really then, it's necessary too difficult to upgrade your fabric, but I mean there are usually like major releases as well. So if you just want to uh upgrade to a major release that can be actually painful.

A

um So these are my my two considerations and what's important actually to note uh that if we speak about like continual Improvement, uh I mean I mean everywhere. If we speak about these management practices again, uh it's it's just just important to to ask ourselves how much it should be decentralized, and sometimes it's not just our choice.

A

uh Sometimes we can configure so like if you have like a chain called deployment, for instance, uh up to a certain point.

A

We can configure how much it should be decentralized but, like you know, if you just want to deploy your infrastructure um It's, usually the good idea, if, if you release your new peers, for instance, for your whole Consortium and not just practically for for one company so overall, if you speak of like like management practices and Organo, if you have a point here, uh the question always raises how much it should be decentralized, uh how much it can be decentralized and how much uh you know like uh how much cross-organizational uh communication uh is, and coordination is needed for a certain anything for a certain something.

A

So that's one thing, uh the other thing that can be considered by continually Improvement. We don't just want to you know, upgrade fabric itself. But again this is like this is like an application which has some other components as well. So it's a pretty good question: uh how we can update fabric with the infrastructure, with the underlying infrastructure and with the with other components as well like with user interface. uh It is, it might be. A very complicated question because, like you know, I mean practically like network uh is, is run by many different companies.

A

uh So you you usually can't coordinate of the individual, individual, continual Improvement processes or individual diploma processes or.

A

Application as well, um another General management practice is, is information, security and in information. Security. I would practically consider two points.

A

Second, one is: is data, consistency and it's a very simple model. I would say, uh because privacy I mean because over like like a couple of categories. First, like in privacy, we got physical privacy, which is which is the channel itself uh or you can use like private data collection, and it gives some physical privacy as well.

A

uh Apart from the fact that your hash is written into the chain which is visible to everybody, uh but that's like physical privacy, we can speak of like cryptographic, privacy, and if you speak, if we speak of cryptographic privacy, then I mean already. There are two categories that should be considered: one is classical privacy uh or classical encryption, and the other one is something which is quantum secure, uh which is still in standardization, but I mean my experience. Is that so usually I mean customers?

A

Ask hey, what's what's going to happen in like the 10 years, with this whole blockchain system? uh If we get, if we need like Quantum secure algorithms, uh can we just I mean is? Is there going to be the possibility to change or do we need a new system and and basically migrate, everything and so on and so on?

A

And we got one category as well: that's privacy by policy, there's no like cryptographic or physical guarantee. uh People are, are not allowed to see certain parts of the system and that's, like you know, I mean hidden by the user interface or something similar.

A

The other thing, that's because that an information, security management, that's like data consistency and again here I mean again the same categories.

A

So we can have accessing, which is a physical constances, so there are like physical guarantees. You know, I mean one channel. Basically, is one big block of of storing something which is not mixed by anything else, so it might be considered as kind of a physical consistency as well. uh We got them blockchain a lot of cryptographic, consistency and it's it's.

A

You know, I mean it's usually the the uh the digital uh signing and the and the hash algorithms for the first round, uh and usually we can consider to be categories as well like like classical or Quantum secure. So, for instance, in hyperledger fabric uh the you use like the ecda, that's the elliptic curve, digital signature algorithm.

A

um You can even consider that are like uh free, free size of uh of key three key size, but that's funny. It's just not Quantum secure. That's for sure, um if you consider like the classical.

A

But like in a cryptographic, hash function, you can just increase the size basically uh or the size of the keys, and then it's gonna be more Quantum circuit. That's not the case like, for instance, by by digital signature, algorithm, it's awkward. So it's a good question are what happens if you need, like you know the possibility for Quantum circular order, it's like in in 10 years.

A

uh What can we say actually for the customers and there's other uh the other topic as well, I mean consistency can be done actually by kind of a policy as well.

A

So you know I mean you get database, which is type write in information and then and then, uh if you guarantee that only the administrators can basically uh access them, then probably it's not such a problem and what's important here uh in terms of Information Security Management, uh which parts are basically supported by by hyperledger fabric, uh so which is like an out of the box functionality and which needs some some development, or or even perhaps, research and development.

A

So these are the uh Information Security Management aspects of high manager fabric and it's again it's sometimes not easy just to set up. Basically, an architecture uh which which fits for the for the customer requirements, uh sometimes even collecting the the customer requirements is, is something which is difficult.

A

So let me just continue uh and that's like more uh so less theoretical aspects, it's again a general management practice which is kind of a relationship, management, supplier management, uh supplier management or or even service level agreement. uh These are actually three different General management practices. I just uh put it everything in one slide so again, I mean here. The question is: what is the level of decentralization in terms of governance, operation and maintenance?

A

uh It might happen actually that there's like one company, maintaining the oil Network so, like you know, I mean I mean I mean even if the even if, like you know at all your your business application, is decentralized. There's there's one selected company in your business business Network, which maintains your network so which, which creates the releases uh which deploys new software, uh which collects the log which monitors the system, and it means basically that's uh from a from a maintenance product perspective.

A

It's it's going to be just much easier, uh more like a classical management uh service management, but it might happen that I mean your. Your maintenance is fully decentralized uh and then and then yeah. So it's it's usually a question um how much coordination effort is required for the different processes?

A

So if you speak about like any points here uh always the question is for a certain point: uh how much coordination effort do we need uh which are the which are the organizations that should be affected and and if there's any support, how they can coordinate uh on chain off chain and so on and so on? And so so uh one thing is always to consider like infinite coordination.

A

How much coordination do we need uh and if there's like a structured way of coordination so like we can see like there's the ad hoc coordination, you can say like we have kind of structured, plus organizational structure uh coordination. We can have like automated coordination, but still independent from from blockchain from hyperledger fabric, and we can have something as unchain automated uh coordination.

A

So, for instance, if you release a new chain code, there's a process or or releasing your your new chain code in fabric version, two: that's that's kind of an uh coordinated, uh automated on-chem way of of delivering a change which is a new business logic in your network um Innovative, but that's for chain code and it can be customized as well. um There can be many activities for that. There's no unchain support, there's, no there's!

A

No even automated support, it should be coordinated, ad hoc and it might, it might have actually a lot of a lot of efforts and one question that I actually I, don't know if it's possible or how it's possible. If there's a way of of defining a service level agreement I mean I, mean I'm sure if you just change back to this topic, yeah sure. So so we can have like, like service level agreement for the individual infrastructure, but especially this part, which is which is hyper larger fabric.

A

This is like you know, more multi-organizational, so I mean I mean there might be like ideas.

A

For instance, and defining kind of a service level agreement for that ordering node, because otherwise the system stops, but it's difficult, so I would say, especially for this part uh defining like service every service service level agreement can be challenging, and the problem is that, basically, this is like core component of your business application. So as a score component, uh I mean for the core component, you can't Define a service level agreement it especially how you can do the same for the whole application.

A

So again, this was like a question. If, if service agree, level agreement can be defined at all in uh in a meaningful way,.

A

To the next slide, that's again it's like the change management, uh so the classical way of of thinking uh life change management. We get like like items, that's called configurable items that can be somehow configured and then and then we just change these elements. So like we got the router, we got the new settings on our router.

A

Then our configurable item is the router uh and if we just I, don't know install a new firewall rule to our router, then that's a change on our route or practically, uh and then so, usually, uh if we have like like containerized application, uh this is like partly automated, uh because uh I mean, if you have like Docker the the park. I mean part of our code. Part of our infrastructure is, is already coded, uh so already programmed.

A

uh So it's just totally different uh different and it is usually not so much considered that we have like, like changeable items or configurable items that should be changed actually actually somewhere.

A

That is subject to change, and then, if you change this item then usually you should consider something which is an impact. So if you get one like like negative rule on all on our on our uh uh firewall, for instance, which is a change then practically it might happen that the whole application goes down, which is actually a pretty pretty serious impact.

A

So the question is how, like such a change, management can be defined in hyperledger, Fabric and again. The question is: if it's like, centralized or decentralized user. Usually it is decentralized, and then the question is how much it is decentralized. uh Usually it requires a multi-organization coordination effort for a change and it's question how much it could be uh actually decentralized or or not.

A

Decentralized and I would say uh it's important practically if you just uh list actually uh make a list on all of our configurable items, all of the items that are that are subject to change and then just identify which are the which are the items that need actually some some multi-organizational coordination effort. So, for instance, uh if you have like your your channel and you want to you want to upgrade. Basically, your channel properties like you, want to have new blocks, not in every two seconds, but in every three seconds it is certainly a changed.

A

Organization and or if you have like several uh several channels, the same channel practically um so that's I mean that's simple or it seem to see that you need to move with your organizational coordination.

A

uh For instance, if, if there's like you know, I, don't know, 11 companies and each company is running running an order order or or let me say we have like free companies and then each company is running like like one one order at node, uh then, basically, if you want to just restart your order order uh at your company, that would require some organizational coordination as well, because if you just uh shut off your your order, a service.

A

Sorry, if your other service is decentralized and you, if you shut off your order, then practically the system hold on. uh So it's something that should be recording. If digital companies as well and again I mean the level of automation is, uh is pretty much questionable if you get like one chain or off chair support or if you get some some other support as well.

A

So the next one is something which is identity and problem management or incident and sorry incident problem, Manager Management.

A

um So usually, it means that something something goes wrong with your system, uh you get an error uh which should be somehow identified, for instance, by monitoring, that's usually kind of an analysis which might which might be a root cause analysis. Then there's some solution that solves somehow this problem. This is usually the change management management.

A

And it's a good idea to to have several database on such incidents so so like. If it happens again, then it will be much easier to solve it's kind of knowledge database on on different incidents.

A

um So that's like incidental problem management and again so it raises several questions: if you, if you, if you want to realize in a decentralized way, so the first question is: how do you want to you know I mean monitor your system. Basically, um you can. You can monitor the individual organizations, uh but then it's gonna be tough to to coordinate and collect the monitoring result or you might as well say hey. We just want dedicated company that monitors everything, uh but then practically you centralize your your your whole operation.

A

But again, the question is analysis that requires some some coordination effort as well. uh We discovered change management in the last slide and look in I mean this knowledge. Knowledge database is usually a good idea. uh If you want to you, do your maintenance or operation a decentralized way, then this knowledge database should be decentralized and build up in a way that like no, no, uh no secular information is shared with other companies, so it might.

A

Looks like Daniel dropped again. If we give him just a minute or so I think he'll be back.

A

So this was my uh 20 seconds break.

A

These 20 seconds week, uh but let me just continue so again, if you, if you have like in independent companies and the independent companies having like its own uh I.T service management, best practices, then it's a little questionable how it is uh you know my whole process can be integrated with them. So let me just go through a little bit more faster, a little bit faster on other aspects as well, so in terms of it Asset Management uh its management. That's usually there, your your components.

A

Like individual components, usually you have something which is a shared component and in hyper natural fabric you don't have like shared components, so shared components is, for instance, if you have a an infrastructure which is which is used by many many different uh companies. So you don't have like the classical shared component components. What you have uh that's I would say it's called replicated components, so these are like independent components, but they are strongly dependent dependent with each other, even if they are physically different.

A

So again, like you got free companies, uh free node ordering service, then of course the one order, and all these runs only at your company, but despite that, that's like very strongly dependent on on other ordering service as well, and it should be maintained and released and upgraded uh together with the other orders or other notes as well, and one big question is: I mean who's responsible uh for, for which part of your of your system or or aggregation that's again, something similar as service level agreement and it it can be actually tough.

A

So in terms of ideas, Asset Management, what's an important thing and it can be actually planned. That's your component life cycle, uh so I would say uh these are the topics that would that should be considered like the the life cycle of of your chain code, uh which is more like. Actually, you know, I mean business oriented, uh so it's the business logic.

A

So basically, business uh will drive somehow the changes of your chain code, but you can consider the life cycle of your fabric, components which are the major and minor operates uh of hyper larger fabric, one thing which is yeah challenging a little bit as well. That's like the life cycle of of cryptographic, elements like at the moment.

A

uh We got like keys and certificates in hyperledger fabric, but perhaps in a in a future we got like decentralized identities and and I, don't know zero knowledge, proofs and other other things as well, so it might be even more complicated in the long run and, of course, I mean one asset that can be considered uh as in the life cycle model. Again, these are the like. The cryptographic, protocols and I especially mean on the quantum thread, which is again not relevant at the moment, but perhaps in in 10 years.

A

uh It's something that should be considered basically and of course last, but last but not least, uh there are some of the components in your application as well, besides hydrological fabric. So let me continue. That's the next one is monitoring and and events management. uh So we got like good monitoring tools. uh Basically for for like containerized applications, uh you can you can very easily integrate with Hyper Azure fabric. It's like Pro, Metals, grafona kibana. With some effort you can even bring like yours up and running and so on.

A

So we got good, tooling support of what's missing, practically uh I. Think I would say these are two things. uh One is uh there's no like Community best practice, what to do with certain logs and math things. So there's a there's, no, like proposition that hey you see this log in your ordering service, then that should be an alert uh and then there should be an email to an operator because then some single divide is going on and if you see the other event or other log, then nothing should be done. Basically.

A

So what is missing a little bit, uh I would say such a community best practice. It's certainly something that can be that can be done on the long run and, of course, I mean. The question is here as well: uh these are tools uh which are which work, which work very good at one company organization.

A

Your organizational stuff here, if we want to do monitoring and even management event management in a decentralized way, so then, how? How do we integrate like the different independent promoters, grafana and kibana stuff, in a decentralized way, with several organizations? How do we collect locks in a decentralized way? For instance, how do we manage alerts, for instance or events or who's who's gets? The email is if something something uh goes, something uh something wrong happens um if we speak of like a fully decentralized or semi-decentralized maintenance operation.

A

So these are hard questions basically or for me at least, then we get something which is like release, management and deployment management.

A

um I spoke, I covered these topics, uh so I'm, just covering very fast again here is I mean my I would consider that we get like chain code, which is like a replication. You might as well have some client components as well, especially that are communicating with with Gateway, uh for instance, these are more like applications, so they're having like the religion, release, management and deployment management idea.

A

Similarly, as an application, apart from the fact that usually we like with your organizational coordination for this one as well and other stuffs, are like beers, CS orders and Coach DBS, these are more like infrastructure, oriented parts and of course the question is I mean we don't want to release hyper Legend fabric. We want to list the application basically, so how do we release?

A

How do we create a deployment and release Management in a way that both Fabric and non-fabric components are covered with that and, of course, last but not least, the question is, uh if there's like five companies each having basically a general release, management and deployment management process that are different than than are different from each other, then how? How do we put actually hyper hyperledger Fabric in the middle and Define a meaningless solution, release and deployment management management process uh in in the middle?

A

So the next one is configuration management, I, just just covered a little bit practical change management, um I would say what's what's important in configuration management, apart from the change management, aspective, that's kind of a cmdb, it's a configuration management database and I mean lately.

A

We got all of our infrastructure or most of our infrastructure practically in code, so this process, or this cmdb is usually overseen but I would say that's kind of an important thing, especially if you have like a big infrastructure, I mean even if, like in in kubernetes, you don't necessarily get everything from from Port history, especially not if you work like with hyperledger fabric, uh because you get a lot of changes which are like you know, I mean if you, if you just change, for instance, the the the the the the the the the the general property like uh increase, the channel timing from two seconds to three seconds.

A

It's not such a change that is, that is Middle or like that is. This is manifested, for instance, in your in your devops environment very easily, so I would say it should be. An important thing to like, like have a history of the of the of the configuration items uh that that's changed uh basically and then it can be like local, shared or replicated CIS, um based on the impact as well, what's practically important to consider that this cmdb partly should be available for other companies.

A

For other organizations as well, because you know I mean for for efficient uh error handling, you just need to see what happened in in in in in the other companies blockchain Network as well, and that might actually resist some some serious privacy consensus as well, and one more thing is, like you know, I mean that's the most user friendly part. That's like service desk. So it's a it's a good question. If you get like uh like a console, two blocks.

A

It's exactly just for hybrid and fabric, then how do you the two easiest approaches or or logical approach? You get the several organizations and then each each organization having like a service desk, probably for other applications as well, so practically each each organization having having its own uh individual and independent service desk. That's that's one way of doing uh such a thing. Another way might be that uh we got like I.

A

Don't know, you know food supply chain, application, which is run by several organizations, but despite we just want to uh Define one interface for the End customer, so we got one service desk practically, even if, uh under the hood, uh there are several organizations um in the network practically. So these are the two easiest examples.

A

uh Things are certainly more complicated if you get like classical service management, uh because usually you don't get like just the service desk and the handy fabric network, uh but you got like the people, uh usually like service desk, that's kind of first of all support there might be like second and third level support as well, so at each stage. Basically, you can decide if you, if you do it like in a in a in a decent life way like this one or a more centralized way like this one.

A

So Sims can be actually pretty messy. uh There can be many combinations and then it's usually I think it's usually the best idea just to consider the possibilities and and choose one because if you don't consider this uh these things and everything is running like in a network way. uh Then yeah, then there's gonna be chaos at the end.

A

So I got one slide, that's conclusions basically, uh and then I finished my presentation. So my conclusion is uh the professional uh Consortium blockchain, like high pressure fabric, operation, maintenance and governance, can be basically challenging uh due to several factors. uh I mean one is I, think the service and application complexity. That's that's. Certainly one issue I mean these systems are, on the one hand, open source, on the other hand, evolving very, very fast, and so they are usually much much more complex than than a very classical applications.

A

Then we got a lot of cryptography and this one is, you know: I mean it's it's moving fast, so I mean the next wave is of zero knowledge, proof, proofs and I. Don't know uh ZK roll ups, uh PR practically probably in systems like in in a couple of years time. Then then these systems gonna be even more complicated and complex for maintaining so there's a very fast moving cryptographic, research and development.

A

Then, what's what's making situation very problematic or very, very difficult is that we get a multi-organization a setup and especially if there's organizations are are Enterprises uh with different standards with different best practices uh with different management, IIT service management practices, then things can be really challenging.

A

uh One thing is that usually there's I mean in most of the blockchain systems. I mean the design goal of of Simplicity or or simple operability is not a design goal. I mean I've I've, never seen a such a system where, like yeah, unless we do a blockchain that can be operated in a or maintained in a simple way. That's that's, usually not a design goal. That's, like you know, priority 20 or something and of course, still there's no industry or Community best practices.

A

uh It might be evolving on the long term, so actually I think as a conclusion, but I mean again, you can disagree with me. I mean it's not a problem for me. uh More Community, best practices and and share experience would be needed for for hyperledger fabric and actually for for other uh for other uh Consortium blockchains or distributed applications as well in terms of like professional operation and maintenance.

A

So that was my presentation uh and then I just finished it. So I'm open for questions or I mean for discussion as well, uh and let me just actually go through on the questions if you have, if you have like question I will try to answer, but in the meantime, if you have, if you have questions, I mean just uh or or comments or or practically you know, I mean I mean best practice. Experience to share I mean you can just unmute yourself and just just go ahead. um So basically it's like yeah.

A

As something which is which is ready and and and Industry best practice, I meant it. Basically, as it's an initial presentation to raise awareness and and then have like more, you know, Community College.

A

Below the YouTube streaming as well and I think I think we're gonna we're gonna email every every participant participants with the slides as well.

A

Yeah, so it's a good question uh how you can get like more information on the on the topic.

A

uh So you can, you can have like ideal experience, basically, okay, so so I I I've chosen, like IIT as one example for um for for profession, professionality service management. uh It might not be the best, but it's my experience that it's it's like better, better in in terms of fabric like uh that another, like you know, very devops approaches. uh But again you can. You can argue with me so, if you're interested in I it, then basically uh you can just browse on the internet.

A

uh There are tons of books, uh a lot of like you know, freely a variable uh you can download them and so on, and so on, uh I covered from itl the management practices which is like more technical oriented or uh not so far from the from the from the technology.

A

But in idea that I, like a lot of uh you, know, I mean holistic, holistic approaches how how your service should be defined, how the value of your service should be defined, how people should cooperate basically in in your organization uh to have like the most value for end customers and so on, and so on? I just didn't cover these things.

A

I'm not quite sure, if there's like you know so there are there.

A

Are there are a lot of meetups, focusing on how to install your hyperledger fabric, for instance, or how to set up in an easy way like uh for a development environment, um I'm I'm, not quite sure how much you can find some some meetups online recording as well for like like planning something more professional as well um yeah yeah, exactly uh I find this as a good example, um I mean I I usually say that, basically, uh if it goes that direction, then then you know I mean I mean such a Consortium, blockchain system or even high blood sugar fabric uh can be as complex on long run, as sap uh I mean at the moment.

A

You don't have so much business logic inside but I mean if you consider that you have a lot of like cryptography as well. So in the long run it might happen that basically I mean I mean fabric is basically a framework with which you can build up many different structures and networks. So, in the long run it might be the it might reach the same complexities as sap uh and then of course, yeah.

A

You need a lot of like you know, best practices, how you can do it easy in an easy way, in a professional way in a more professional way and, of course, yeah. It's it's a lot of complexity.

A

So again, uh if you have like anything uh any questions, any ideas or or best practices, or even if you I mean disagree with me, I mean it's not a problem, uh just just go ahead, uh I mean use the chat or, or you can unmute yourself and just just go ahead and ask him.

A

So, if um I'm just reading in the comments, if you want to, if you I mean if you don't have experience with hyperledger fabric, then um of course my my presentation was practic, not necessarily very useful for you, but so if you want to run your first Network in a development mode, uh there are a lot of good articles, uh basically a lot of good tutorials. So it's not so challenging.

A

uh If you plan to go, live with like 5-10 organizations and then maintain for the next 20 years in a professional way, uh that's that's the way uh that I was that I was more like considering in this in this presentation.

A

So if there's, if there's no more question uh or or anything that you would share, then I would say I would thank you very much for.

A

uh And then this presentation will be available on YouTube again the slides gonna, be your participation and attention and then well um see you see you next time in the in the next Meetup.

A

A

Foreign foreign.