 Hi everyone, can you hear me? Cool, awesome. Okay, so thank you for joining in today. I'm Malolita Sharma and again we have all met virtually I think more or less but it's so nice to see you all. Welcome to OpenTelemetry Day. We are going to be doing a community day today and again we have many of our maintainers here, many of our contributors again everybody who has been participating on the project as well as all our observability groupies I would like to say but you know we're all here. Okay I think we're too close to this. Might be better? Okay so I'm Malolita Sharma and Morgan McLean who was going to be co-presenting with me is presenting virtually in kind and he couldn't join us but I think he did send in some clips which I'll try to play and of course the final recording will be with some of his parts added in but we kind of improvised at the last minute so bear with us if you know things are not perfect. A little bit about myself. I'm member of the governance committee and a co-chair for the CNCF TAG observability. I'm also an observability leader at a major company that I have not put here and senior manager. I was previously a senior manager at AWS working on OpenTelemetry, Prometheus and Grafana very you know again a high AWS team happy to have you here and of course introducing Morgan also. Morgan is the director of PM at Splunk. He's also a fellow OpenTelemetry GC member and previously a PM at Google working on observability tooling and has worked also at Microsoft in the past developing web services. With that said I'm going to kind of switch over to trying this out. I think he did an intro but I'm not sure we'll see. Hi I'm Josh McDonald member of the OpenTelemetry technical committee and an engineer at Lightstep. I'll be talking about metrics instruments and the requirements that gave us the up-down counter a new kind of instrument. Let's review how we got here. Metrics systems have been around a long time but I'm thinking about what happened around 10 years ago. Looking at open source systems prior to then we had just one kind of metric instrument. I'll call it the number instrument. We would use it to report numbers with timestamps and then we would plot those numbers as a function of time. The result is numbers in a line. The nice thing about numbers is they support math. You can do a lot with numbers but first you should know what they mean and 10 years ago the interfaces we used to report metrics began to change. Why? Because it helps the metric system help the user when numbers have meaning. It's said that you can just hit three. We have our audio back. We were kind of expecting this because they were doing this impromptu and the gentleman he just said press four first and then press three. Going back to our regular programming. I'm going to talk a little bit about the milestones that the project has embarked on since 2019. As many of you know open telemetry evolved and has been the output from two big projects you know converging open census as well as open tracing and in May 2019 open telemetry was announced as the merge project and with the mandate to provide APIs, SDKs and an agent for metrics and traces as well as converging the land of tracing and metrics into one project with the communities working together across the board. Fast forward to April 2020. We began tracing RC and GA releases. We began work you know shifting into instrumentation and metrics. 2020 May logs was added as a new signal which is pretty significant. So now you know this is back in the past two years ago and as you know many of you who are involved in the project logs has been continuing on as the project has kind of focused in on tracing and metrics. Tracing landed in as RC and then GA last year in September 2020 and in January 2021 we embarked upon an ambitious project of you know meeting with the open metrics and the Prometheus communities and then kind of figuring out how we needed to overall are and enhance the metrics specification which guides the implementation of metrics on the project and then there was a significant amount of work done which has you know been ongoing through the Prometheus workgroup on the project as well as the work that has been done by the different states on the different languages 11 languages are supported on the project and to enhance metrics implementation for the collector agent as well as the APIs and SDKs in 11 languages. So fast forward to December 2021 so let me just make sure that you can see it. So September 2021 tracing was finally completed as GA and the stands are logging agent which is kind of one of the foundation components of the logging work that's happening on the project were added right so big milestones. December 2021 the logging data model was finally and the protocol for phase one have reached GA so that's very exciting because you know as the project completes metrics implementation will switch over to supporting logging and May 2022 in at KubeCon in Valencia the project announced metrics GA and RC releases for different languages and profiles got picked up as being added as a new signal. So fast forward to what's coming up on the project in the second half of 2022 as well as you know going forward into 2023 is the completion of logging and we are expecting that logging releases and implementation will go in for RCs later this year and then GA is targeted for the first half of 2023. So that kind of gives you a glimpse of you know some of the milestones that the project has hit continuously and again as many of you know milestones are also dependent on you know the number of engineers contributing to the project at any given point in time so if any of you are not contributing to the project yet please do join in because that helps us you know get get to stability faster. So the current status of the open telemetry signals just wanted to go into what is you know what does stability mean here. So open telemetry looks at stability with these different areas being supported for the three major signal types traces metrics and logs and as you can see it's a requirement for the project to have a well fleshed out data model and a protocol before an implementation for the API and SDK is completed and as you can see these are the dates when tracing has hit GA metrics we have been rolling out gradually and I'll go into more details on the breakdown of where we are with the different metrics implementations as you can see in the last you know part there are five RCs in progress and have already been released and seven are in progress so we are anticipating a rolling GA for the different languages you know through this year and then we have logs as the final signal which is in progress right now as you can see the protocol has already been stabilized and again API specification as well as SDK specification is the work in progress again there's a few who are interested in logging please feel free to join in for those discussions there's a lot of work to be done on metrics also so that's something that you know again in the implementation area would welcome anyone who is interested in that area to join in as you can also see logs and in terms of implementations are not done yet that's something that will actually start rolling out as RCs later this year and then following into the next year so what's the current artifact status this is you know for the different languages the again as I said open telemetry supports 11 languages and APIs and SDKs and 11 languages these are the top ones you know which have the most amount of activity and then there are others which are not on this list but you know are in progress also as you can see the collector which is used as an standalone agent is in GA for traces beta for the user APIs it's in RC the user APIs is stable at this point so you should be able to you know take it and run with it as an end user there are some improvements and enhancements that are being made for developer APIs and that's you know work in progress Java is in RC for metrics JS is also in RC Python is in RC and as of the late as of this week C++ as well as .NET have already hit GA and of course as you can see go Ruby and PHP are close close almost close but they're not at RC status there's work in progress again for any of these languages you know if you're interested please get involved and help and then comes logs right so there's a lot of work to be done and at this point you know there's a lot of work happening both on the specification as well as on the collector in terms of you know an alpha implementation and then as the APIs and the SDK spec matures you know implementation will happen right after okay so switching to community right so I mean many of you know as you've all been involved that open telemetry has seen an amazing birth in the last three years and you can see you know literally more almost 1,700 companies participating in the project it's really phenomenal to see the kind of birth that the project has seen open telemetry is the most popular project on in observability under the CNCF right after Kubernetes as you can see the top top row is for Kubernetes with 761 K CNCF reports on GitHub activity and then open telemetry is right after that 254 K which is which is pretty phenomenal it's amazing with the kind of energy and the collaboration across the different projects and teams that this contribution has been possible so far and we still continue to see it growing and maintaining momentum as we go into logs which is in one sense one of the most well established forms of monitoring as you know but at the same time also there's an enormous amount of complexity in implementing logs well just wanted to do a call out most of the world you know that revolves around observability is involved in this project some of the biggest you know vendors who are very much involved in building out observability solutions or services of their own are involved as well as you know just open source contributors who have formed the fiber of the project as maintainers as well as leaders on the project again huge thanks to our contributors you can see independent is a very interesting you know call out here there are a lot of inter independent contributors to the project you know folks who have actually worked across different teams or work independently as developers and it's pretty cool to see that you know strain of independent contributors also on the project which is you know pure open source contributors there are also a lot of end users in the involved in the project who give feedback who are part of the SIGs who have been discussing and providing requirements as you can see you know many of these companies who are listed here and organizations are very involved and and we continue to see more growth on you know end users actually coming in and joining the project to provide requirements and being able to actually work with you know voicing some of the requirements that they have and you you know configuration management for example is a big area I also wanted to call out some of the extraordinary collaboration and interop work that the project you know open telemetry has done and has been involved in and continues to be involved in again huge thanks to all the contributors from these projects who have worked very closely with us many of us are close friends close collaborators you know and and really love of the ability and open source so as you can see Kubernetes open metrics Prometheus Jaeger, Zipkin, Pixie, Cortex, Thanos, Fluent Bit, Fluent D, StatsD, Prom Scale are some of the big projects who have been involved and worked closely with us we continue to have work groups in these areas where we have discussed everything from protocols to what makes sense from an observability standpoint API's looking at compliance tests again we worked very closely with Richard is there from open metrics as well as Prometheus you know Prometheus project to drive some of the open the Prometheus compliance tests you know which were announced by the project and very proud to see the collaboration in terms of compliance tests actually being part of open source projects to you know ensure that the end user experience is really first class and there is full compatibility for consumption of observability data across these you know platforms so what's next on the roadmap and you can see there's lots of activity just happening above and beyond just the implementation of the APIs the SDKs and and and the foundations if you will of supporting you know the APIs and SDKs and the collector agents of in the languages and those are semantic conventions as many of you know semantic conventions are very important for being able to define and support correlation and insights that are gathered from different types of data signals and so that there's a fair bunch of work happening around there there's a work group that's that's being used then there is also and really you know very helpful but very useful end user feedback group again some of our folks here Char and others are here who have been you know driving this effort Austin Ted Young and and again getting feedback directly from the customers who are actually super interested or have deployed open telemetry for as a collection solution for their infrastructure for their applications and getting feedback directly to the project for being able to you know tune the functionality which is really cool because as you know open source you know is as compelling as what end users can do with it and and what developers can do with it so this having this kind of a work group and an active dialogue really helps another area which is been supported and and was added to the project last year is client telemetry and this is a very important area as you know many of us come from the server infrastructure world where server telemetry in observability is for you know quite well defined and supported but client telemetry is equally important for those of us who have networks of devices or iot applications or web browsers or programs you know that are running on browsers it's it's equally important to understand the observability and observe the signals for system performance and system health end to end and and so we look out for you know the work that's happening in the client telemetry area which will then again intersect as a stack with server observability to actually provide a full stack of you know really system performance system health for those networks that actually you know support end to end not only services another area a set of areas which is very active and also we have work groups for these are client instrumentation as many of you know instrumentation is and very important part of enabling easy adoption for end users and auto instrumentation especially is an area that the project has been working on as you know java is fully auto instrumented another area that's another language that's coming in as javascript and i know are some of our maintainers will be you know covering that later today so again you know we are the objective of the project is to make instrumentation as seamless as possible and and again we invite you know any collaborators who are interested in improving instrumentation to work with us extending observability all the way from client apps and web apps you know taking spans metrics metadata logs and making that available easily another couple of work groups that have just kicked off in the last few months one is agent management and this really you know again goes back to how do you make agent management easy this is a complex area because you not only have local you know configurations for like an collector agent for example but you also have remote configuration right for many large scale networks that end users have remote management is a key part of you know how they isolate across their networks and at the same time are able to interoperate with configurations and allowing sdk agents and the collector to have their configuration remotely updated is a big big is a big area of work that's ongoing there's a design spec which is being discussed right now on the project especially in the collector's sake again if you're interested in you know kind of even tracking the instance of system health for the collector agents that are running on your services or on the end user networks understanding health and being able to communicate status is something very interesting and and you know again this is the work group that's working on this similarly again and this is amazing you know this is how open source works we just kick started a profiling sake where we are starting to look at how we can actually introduce and add a new signal that will bring in visibility out of the box for code performance and this is just started from you know discussions with the larger community and maintainers at kubecon earlier last month and there were a bunch of discussions before that because we've been working with pixie and other projects very closely on you know how profiling could be supported but there's a new sake that we are that we have and and again needless to say if you're looking to participate in any of these areas there is a calendar that you know for the project that you can just sign up for and just join any sig or work group that you want to that's interesting to you you can attend some of the meetings while there's specific discussions ongoing there the project does record and maintain you know very good notes about each of these sig areas so you can catch up on them later however you would like to see better tagging of our youtube videos but on the other hand we do have them recorded if you want to give and catch up so just wanted to call out some of the major you know areas that we are working on and and of course you know going forward then we also many other areas which I'm sure will be you know can easily spin up work groups on and and you know group of folks who are interested in that area just work together okay so again you know with the thousands of contributors who have contributed to open telemetry there is many there are many ways to get involved and needless to say everyone is welcome I really have found the project to be you know I've been involved with the project for three years and it's really an amazing project to be involved in there's just so much activity and so much discussion and and really really nice people that you know we collaborate and work with so so again I welcome you to join in as a developer again who is not an open source pro but is getting involved in open source from their company or from your organization again it's very easy to join in you can just listen in lurk on the sigs get involved work on the read the specification there's a lot of documentation on the project open telemetry.io is our website so you can easily you know go and browse and of course our code is on github I also have the link there but again feel free to you know get involved in whatever way you can code reviews design reviews contributing bug fixes new features tests we have you know contributing built out CICD pipelines continuous security pipelines release tooling again support and you know contribution in 11 language spec areas so needless to say there's a lot to be done right and and you can easily get involved in any of the areas if you like to write features get involved welcome you to do so if you like to you know work on infrastructure CICD you see a continuous security is an area you can get involved in and and also would like to call out that open source projects of this size and magnitude are usually not single-dimensional it's not only developers who are working on these projects we have really an extraordinary number and and wonderful community or product managers community managers who are involved in the project again you know we all work together brainstorm about you know how best designs can be built in for the different components and you know again open source iterates and and polishes and improves continuously so would welcome again you know there's lots of opportunity to create awareness to actually educate operators and end user developers to use open telemetry well to add project documentation to share knowledge you know work on semantic conventions if that's your thing um run an open telemetry meetup I think we have a really nice community across the world of folks who are doing online podcasts as well as open telemetry meetups in person publish blog posts and demos on the project site you can contribute it's as simple as doing a PR and also welcoming end users you know to again join in provide feedback you know your feedback matters a lot get your developers involved in open telemetry provide your use cases and requirements and these are just some of the areas you could you know there's a lot of lot of energy and enthusiasm around these last but not least I'd like to thank all of our contributors without whom we wouldn't have such an awesome project our maintainers our responses and host again very thankful to CNCF for you know actually pulling together and providing us the support for the for doing open telemetry community day today and thanks to our sponsors without whom we wouldn't be able to pull everybody together too with that said again have a wonderful rest of the day and again welcome Austin to come back and he's our MC for the thanks