 Okay, thank you. You can proceed. So in terms of general logistics, we have the hotel sorted now. It's this family hotel in Kigali. You can Google it. It's pretty much in the center of town. Running from the 27 to 30 of March, the dates are fixed. It's four days, Monday to Thursday. It would be kind of nice to do a fifth day in the sense that all the topics that could be discussed, we end up kind of running a bit short on time. At the moment, that's how we are stuck on the four days. Coming immediately after the week before, there's an interoperability workshop taking place in the same venue. So, general program running from eight in the morning through to five in the afternoon. What we do on the first morning, a little bit of introduction and discussion of the program. We'll go through the program with participants. Introduce ourselves to one another. Because we plan to do quite a lot of practical exercises, what we've done before, it can work fairly well, so we'll do it again, is we'll provide a cloud server to each pair of participants. It's not one each, but between two people. And they'll have that machine for the duration of the week. They'll have full control over it. They can destroy it, rebuild it, do everything from scratch. And so, it takes a little bit of time to figure it out. Logistics, picture, where everybody's got their accounts created that they can log in and stuff like that. But hopefully by 10 o'clock, we'll be done with all the logistics and introductions. And then we get started into the meet. The first session would be just a little introduction to the architecture of DHS to implementation. The way we typically do, the Tito will probably run this, I think. We'll do an install, you've seen the install, but the emphasis here is not on how to install it, but really looking at all the pieces and understanding how they fit together and what's required for a typical DHS to implementation. And we have lunch. And then the first thing I do in the afternoon, we'll talk a little bit about SSH. I know we always hope that the participants, when they arrive, they're already kind of SSH wizards, but the reality is, in most cases, there are a few people who are not, and others need a bit of refreshment. So, yeah, just to confirm everybody's happy with using SSH keys in particular, but also a little bit around SSH tunneling, jump hosts and a few things like that. If any of you have attended these things before, you'd be familiar with this. I think I did an online presentation of this last year as well, a year before. Then in the afternoon, we hope we're going to be joined by Michael. Some of you might have been on the poll, I don't know, three weeks back, where we had our security lead, Michael Markovich, present a little bit from his side, some of the work going on from the security team. I hope we will have him there in Kigali with us, and he can talk a little bit about security of the reference implementation and things like that. What I try to do at the end of each day, we should have a quick 15 minutes, just to review how the day went. It might be that people feel certain topics have been covered adequately, or they may want to find time to go over something again in more detail, whatever it is. So, at the end of each day, we have a little after action valuation of how the day went. And I see this has been added, we have a cocktail reception, that's nice. Okay, didn't know that, that's good. And then we're all going, you have a cocktail reception. I suppose this is a get-in-the-witch kind of thing, looking forward to that. Day two, start the day in just 10 minutes. Just quickly discuss what my head is going to be. Again, find this useful to do just to make sure everybody's on board with what's the topics that they are. We're going to spend a session talking about LXD and containers. It's kind of a little bit opinionated. It's kind of specific to a specific type of installation, but it's the kind of foundation to most of the practical work that people are going to do in this. So it's quite important to just get a little bit of overview of the basics of LXD and containers followed by some practical exercises, creating containers, setting up storage pools, setting up networks, that kind of thing. Then we start digging in the details of different parts. So we're going to spend a good while just looking at UMCAP, right? I know that's not the only container. There are some people in the world using Jetty. We're going to focus on UMCAP. A lot around securing it, a little bit about tuning it. Yeah, all the important pieces about UMCAP. How does it log in? How does that work with the DHS too? Followed by, now this is something we haven't done really properly before, and I think it's long overdue. The DHS comp file has grown over the years, and most people typically only use two or three options in there. Dito, are you watching the waiting room? I can see people occasionally coming in. Yes, I'm actually actively watching and letting them in. What we're going to do here, there is some reference documentation on the DHS comp file. I think it's actually not as complete as it could be. Some of the options could be explained much better than they are. As of having this session, it will be good I think in terms of improving the documentation as well. But the other good thing, I think we have Austin is going to join us. Austin McGee, who's DHS due technical lead, deputy technical lead at the moment. And we may give this session over to him. Go through all of the configuration settings. I think that'll be a very useful and informative session. Phew, then we'll take a break, have a cup of tea, and we'll have a session on the database Postgres, how to install it, some of the particular configuration settings, which are particularly DHS to a little bit around, around getting the maximum performance out of Postgres, getting a bit about logging. And I'll reach the end of End of Day 2, quickly after action evaluation. No cocktail reception again. We only get one, it seems. Then we move on to Day 3. Day 3, basically continuing with picking through the pieces. We've looked at Tomcat now, we've looked at Postgres. We've looked a little bit more on Postgres, particularly around the topic of backups. How to make a backup strategy, backup plan, how to activate that backup plan, some of the particularities and complexities around backing up DHS to in particular, ways of using alternative ways of doing offsite backups, S3 storage, things like that. And again, followed by a bit of practical exercise around some of those ideas. Then, nearly there, looking at the other important part of DHS2 installation, which is the web proxy. Now, as people who know me well know, I'm a big fan of Apache 2, I've never been very fond of Nginx, but I know that, I mean a minority on that. And probably we'll base that session mostly around Nginx, in which case someone else will do it, other than me. We might, as the time goes, might squeeze it, maybe I'll do 20 minutes talking about, talking about using Apache 2 as a web proxy. Either way, it's not fully decided yet how we run that, but either way, we can now look at the proxy and just expect a particular configuration. Then we have lunch and then it always gonna tell us a load of stuff around Ansible, automating the installation. You'll have already seen the Ansible automated install, because probably I've demoed this on day one, but going into a little bit more detail, what's inside all of those roles and how you might customize and where the points are to extend it, how to make your own, that kind of thing. Again, we'll generate some exercises around that. And then just to break things up, so it's not all just death by PowerPoint. We have a bit of a discussion session in the afternoon. This is really an opportunity for the participants to talk about some of their own hosting arrangements. I'm always very aware of these kind of things that there isn't a big separation between the facilitators on the one hand and the participants on the other. I mean, we're all professionals engaged in the same kind of business. People have different expertise, and certainly don't expect that all the expertise is coming from the organizers side. So yeah, I hope we have a good discussion there about some of the different types of installations that participants are using themselves. After action, evaluation, and on to the final game topic. I think Emion referred to this a little bit last week. We've not enough time, we've allocated here. We're talking a little bit about both alerting and monitoring. There are two quite big topics in themselves. How to set up email alerting and then alternative ways you can do alerting. Again, with some practical exercise. We have a session planned on troubleshooting. Troubleshooting is really important. If troubleshoot, it's a lot of DHS2 instances over the years. I'll try to synthesize some of the things I've learned, but also we'll hear a bit from everybody else who everybody's got. Particularly people like Gerald have been doing DHS2 for a long time. You'll have solved a lot of problems over the years. Let's share experiences and set up or fixing things when they don't work the way we expect them to. Some discussion after lunch about future architecture. Again, we've got Austin joining us, which is good. I know he's got lots of interesting plans where DHS2 may be going over the coming years. We'll talk a little bit more about using DHS2 with Docker. Kubernetes, way DHS2 might be moving more in the direction of microservices. Hoping for an interesting presentation and discussion there. Then we'll finish off in the afternoon, talking about probably the most important session of the week. Doing a short week like this, you can kind of rapidly go through lots of topics. The real learning happens in the field as we go through the process of trying to manage this difficult beast of DHS2. How do we continue that process beyond this four-day workshop? Discussion about how we want to organize best around our community of practice. We've tried various things over the years. We started off with a mailing list, a Google group that didn't actually work very well. We started with Lameen, in fact, started the Telegram group that some of you are part of. That proved to be very popular. There are some weaknesses with the Telegram approach that I think part of the problem is it's not very transparent. It's not maybe the best way of accumulating wisdom that people can go back and search through and the like. We do of course have the community of practice now, the community of practice website that is. The community of practice is just ourselves. But yeah, let's have a discussion about how we want to best build this community going forwards. That will get us to the end. So that is the plan. It's going to be in Rwanda. So we expect we'll have a load of Rwandans. I see Hamza is there. Hi, Hamza. We're looking forward to coming to Kikali. If people want to scroll through this thing, I wonder if I can actually share a link and allow comments. Let's try something like anyone with the link. Okay, I'm going to say anyone with the link can comment. How about that? I'm going to copy the link. I hope I don't get into trouble doing this. Where is the Slack chat? Within the chat. If anybody wants to go in there, make any comments, scroll through it themselves. They can do some. Any time. Any immediate feedback, any thoughts? How does it look? Is there any things you think we've missed? Do we need to take an extra week and do this over? We could do a month of server admin academy. Unfortunately, we just have the four days. Can I go? Yeah, we see a schedule. Yeah. Can we have some session about load balancing? About load balancing. Yeah. Load balancing on the Tomcat side or load balancing in the back end database. With the boat. Work will be great. Yeah, we've got this kind of really. Small session here on scaling. Scaling in there. It's probably not sufficient. We've got something here on. Yeah. Okay. We'll find a way to fit it. Something we'll have to give away, but I know it does come up. Often. People having performance problems. So they want to run four Tomcats instead of one. And usually they haven't looked at what the performance problem is. And the problem usually isn't solved by having four Tomcats because the problem in almost all cases is at the database. That's why I have not really emphasized it. But yeah, we'll find a place to put it in. It was on the back of my mind anyway. Can you make comment on the document? Find a place to put it. I mean, it's going to come up there. When we look at Tomcat, it's going to come up there. We'll look at the file. Maybe we can find another way of. It would be quite good to set up some kind of practical implementation of it and do some performance testing and stuff like that. Make the comment on the document and I'll have it. I think maybe a section on the future. Ideal setups for servers as well. So the Docker and stuff, it's becoming, I think, more popular at the cloud. We have a session planned on that, I think. Okay, sorry. Right at the end, future architecture considerations, this one. Oh, yeah. Sorry, Bob, as I said. And what I'll probably do, I'll try and persuade Austin to do this. It might be adding on the backup, your best practice to restore and resolve from a critical outage, disaster recovery, maybe. Yeah, it's about having a backup plan. You don't really have a backup plan unless you're testing your backup plan. You know, we'll have a backup and let your test restore your backups. Yeah, we'll definitely emphasize that in this session. Okay. Sorry, somebody else was saying something about the money toss. I don't know if we can have what you commissioned about setting up for me curious. I'd love to do that. Kind of. We really need an extra week. Because we also want to be nice to talk about Zabek's Prometheus, the Elk stack. We've got a very brief section here on, as I was saying, there's kind of too much squeezed in here to an hour talking about monitoring. I mean, if we were going to do two ways of doing this, we can either have very general presentation, all the different kinds of monitoring, different kind of tools out there for monitoring, including Prometheus, Grafana, et cetera. Or we can take an example and push it through to the end. Or we can all we can dig into some of the slots that we've got free for practical exercises. I don't know, Dajo. There's not actually too many people using Prometheus for monitoring DHS to be interesting to hear from those who are. Damian, what are you guys using for monitoring your local containers? Nothing at the moment for monitoring. We tried Prometheus a couple of years ago, but we had the basics set up up and running. How key DevOps, IT guy left during the temperamentation, so it's still lingering. Yeah, that's typically what I've seen with Prometheus set up, Damian. It's quite easy to set it up and then you face with quite a big job then of actually turning it into something really useful. Most organizations kind of run out of energy at that point. Interesting to get. I mean, the fuck I know we are using Prometheus is GAO. Could maybe get some input from them. Let's talk to our own DevOps team. Phil and Rada one then. It might be a good time where we can come and have a bit more time or extended kind of workshop and you and your team to go through. Sorry, I missed that. Someone's talking about more extended time. Yeah, I think he muted himself or themselves. Who was that? Sorry, I miss that last comment. All I heard was something about more extended time. I think we have to realize that during this one week get together, we've got to have fairly realistic expectations that all the stuff that we'd be able to touch. But the important thing is that we establish good mechanisms around building this community of practice building because most of the learning and experience sharing and the like is more likely to happen during our weekly sessions like this than we're going to squeeze in to the week in Kigali. Frank, what I'm saying, can we include upgrading a DHS2 instance? Yeah, that's a very important topic of upgrading. We haven't actually put it here. Scheduling upgrades, how to actually do them, the difference between patch upgrades, major version upgrades. Yeah, let me make a comment to myself in the document. I'll put it here. It's not necessarily here so I have a comment. How do I enter the comment? Is it a room? Why can't I go into my own document? I'm missing something. Ah, I had a comment. Just to make sure we cover this. No, the latest talk from the release engineering team is that we're planning to move to a cycle of annual releases, move away from currently we're going to release, doing a major release twice a year. We want to move to doing a major release only once a year. I think that's going to make life a little bit easier for a lot of the implementations who struggle on the treadmill of trying to keep up because, you know, we only have security support for the last three versions. But yeah, some talk about the cadence, the cycle of upgrades. We'll fit it in some way. Good point, Frank. Automatic backup from Gerald. Yeah, we're going to talk all about backups, everything to do with backups. When you say automatic backup to the locals, I presume you're thinking of bringing a backup of database from a cloud instance and bringing it back in country. So having an offsite backup. But yeah, we'll talk about that. Here are the comments, thoughts. Tuzum, you've been quiet. What topic would you want to see? Yeah, sure. I'm actually following up. I'm actually following up and yeah, it's good. It's a good plan. And when it comes to addition of comments, I think I will just pop in. But as of now I think it's quite comprehensive. Yeah. I'm painfully aware Tuzum, because we won't be able to cover everything. Or if we try to cover everything, then we'll do it in a very superficial way. So it's trying to find a balance. Yeah, feel free to comment on the document anyway. Yeah, yeah, sure. In terms of resources, besides myself and Tito, as I said, we are hoping to have Michael on the security team and Austin and possibly even somebody from the UIO DevOps team as well. Some of you met Radom the other week. Could be Radom or Phil or somebody else. In terms of entertainment, we have the cocktail. Usually we try and do a dinner as well. So we'll have dinner together, maybe on Wednesday. I have to speak to Alice, what the entertainment budget is. Okay, so for those of you who haven't signed up yet and registered, if you're interested, then please go register. Because I know that time is going to be a little bit tight in terms of organizing funding and the like. But I'm looking forward to getting together. It's quite a long time since we've done this. I think over the last time we did this was actually in Togo, in French. That was hard for me. I think Dajo did most of the talking. I was just an observer. We might do it again later in the year. I know, for example, there's quite a few folk from Vietnam, but it's probably easier. You may find it quite hard to get here. So there has been talk about doing another one, perhaps somewhere in Asia. But we'll see how that goes. It'll be the guinea pigs. Okay, well that's it from me anyway. Tito, I don't know if there's anything you want to add. Okay, yeah. So I think the document is very comprehensive. We have a lot of topics that we're going to discuss about. And guys, feel free to add more comments on things that you want us to talk about. So we are preparing the presentations and stuff that we're going to go through during the academy itself and even more practical use cases. So I think it's very comprehensive one that covers a lot. And as Bob mentioned, the time is limited. We only have four days, but we have a lot to cover. So yeah. Yeah, we're trying to do it also as kind of as sort of a general as possible as well. So, for example, when we're looking at in detail at Ongat, this shouldn't really make a difference, not much of a difference, whether that's running inside a Docker container or an LXD container or even running on its own virtual machine. We're going to look very specifically at the application itself. So hopefully we'll manage to meet as many people's expectations as we can. I think we've got quite a long session. We've got nearly two hours looking at Ongat. We've probably got a bit of time to get into some of the load balancing discussion there as well. I see Damian talking about use cases and pre-live examples to compare and analyse. Also, maybe it's another comment. Yeah, I think the only misgiving I have looking on this. We've already really got one session here where we're going to involve the participants a lot. This, where is it? Where is our discussion about infrastructure? Down here on Wednesday. Other than this, there's a lot of stuff coming from outside, not so much coming from the participants. We might try to find a way of either bringing forward a program a little bit or trying to find some other way of involving participants more. One of the things that we've done in the past with these practicals is that we provide an opportunity at some stage in the week where participants can present what they've done in their practical exercises to the rest of the group. We might try to build in some sessions like that. All right. I'm just looking in the chat here. Just going to real-life examples to compare and analyse and be great to review. Yeah, this is kind of where I'm going to with this, Damian. Getting people to talk about their real implementations. I'm not sure if we're giving them enough opportunity to do that. I'm going to try and find a way to skew the balance a bit more in that direction. I've probably realised that more participants are becoming from Ministries of Health than from NGOs like MSF. Maybe that's a better presentation to do primarily and other organisations or setups can be secondary to that. Yeah. Now, we're definitely going to try to find a way to get participants to talk very much about their own setups and have some good discussion around that. I know a lot of that will happen in the evening and over the tea breaks and things anyway. The programme is a little bit packed in. In the past, so I've managed to keep two or three slots open to be decided because during the course of the week you often come up with topics people really want to discuss. Then you've got to find a space for them. A bit more crowded this time. With that point taken, Damian. That's what we want. Okay, there's no comment people want to make at this point. I think we can finish a little bit early. I don't know if I want to put that link under the community of practice and have the whole world coming in making comments on the document. I suppose it doesn't do any harm. Maybe share a local version of it or something. Yeah, I don't know. The link is out now soon. I have more feedback we get on it, the better, I guess. All right, so that's it from me. Okay. I guess we can end it early today. And the document is still open for comments. We've got lots of work to do in material together on these things. Yeah. Okay. Good. Thanks, Tito. All right. Thanks for all of your joining. Thanks Bob. Thanks Tito. Cheers. Thanks very much. Thanks Tito.