Bestand wählen
Merken

Chef Journey on Google Cloud - July 12, 2016

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
I guess or started here thanks so much for for coming hour session you end up on the 5th floor replace 2nd only to the 3rd floor or in terms of people going to the wrong places so thanks binaries J. Marshall Hôpital strategic advisor on who call platform um joined by a whole crew here's literally introduce themselves as we go but as you of a different session here today how many folks here have actually logged in and tried will call platform all that's awesome 2nd is give you 45 minutes back if you want if you already from where the platform and also is a good thing then you're here have you tried a platform on decided to show up I worried do this little bit different way I think to Brandon's comments earlier on this morning the keynote when the big things were really
trying to drive home about GCP is how work have approaching the cloud computing market a little bit differently than they would a lot of folks to it's really about the open API is open
source so and so forth and also a marketing at the bottom here says that the customers users because they love us but we did talk about how much this week so hard love at all can go the other and that is also very true I think it will we consider to be our responsibility to prove ourselves as not only price-performance leader but also ease of use quality the infrastructure so and so forth so do you 1 quick example now you may see a space appear my thing OK will how you supporting something like in age space I made you might be familiar with the fact that h base was born off of the Big Table white paper that Google wrote back in the mid 2 thousands so when we decide actually externalize big table and start letting other people rented offer was essentially we can actually expose the internal plumbing we said look there's a whole ecosystem out here on a space so we actually used atabase API to front end our Google big table service this is 1 example to were nineties obviously another good example of course down here in the lower left-hand corner where
ships so although we hadn't been different add recipes in different you know the knife plugin for dual GCPs so and so forth we really been doing a lot more focusing on the last 6 to 12 months around how Google engineering chef engineering and the partner ecosystem how were all collaborating is a lot more love a lot more hugs from to shift in the chef ecosystem so so I show so since have you know a good food metaphor nary
soldier chef graph and we're approaching this kind of argument and a foodie prior no big surprise there but so they they watch like top Shafran cooking shows you do things like something 3 ways right surrender Shaphan Google 3 ways so far they can wrap or tenderloin is gonna be a chef Ottoman version of delivery were shift work like a confuse I'm sorry just yesterday I haven't finished speaking in India and it happens so ab engine was kind of Our foray into cloud computing right we actually start with engine back into the house and 7 2008 we just kind of wrong brain you'll wanted virtual machines network so so have tensions been around for about you 8 9 years I favor you use a snapshot of the kids use natural that runs on map engine right very popular very large Platform a Service offerings so from chef the lovely JJ and the also lovely thousand from Google after the engineers actually working on that and I give you a demo and show you what we've done in collaboration on there the and then for Ferrara can in my personal favorite the whereas the dual the word the diver view on once again why does dual followed an our platform differentiation matter to shaft or what do we think it should matter and finally for a 3rd course are bacon and tomato passes as actually be as working with our partners for 1 0 shipper Cascadia working on a Multi-Cloud scenario of where maybe you want to use not only cloud other top providers and maybe a private data center on a real-world customer example of being collaborate across multiple times because once again us winning doesn't mean other clouds lose right everybody we talk to has a multiclass strategy should I ask how do we become part of the multiclass strategy so that being the
case all terrorists of the guys really care about so Nelson JJ issues yourselves were look checked livery on App Engine my the other thing so I'm not only German developer that will and I've been collaborating
with shaft forest reputed is them so did you the idea here is we're going to use chef delivery shoe rollout and application date and make sure that's as developers usually do they make mistakes and talking of Pacific and I will explain mistakes but we don't want to the infrastructure to suffer as the happen right we don't we want to make sure that we have a system pipeline delivery III and that we want to that the a longer with the delivery so so be end-to-end them all for all all the way from have mapped it run all the way to have my application for the go and that they might didn't also use J. does all manner of our own very I after the city
dealt they OK so I we have we set up a requested or sort of make us on the call delivery on when we set up a delivery cluster and this is our our application and we've created a did have repo with our application which is you see is a pretty straightforward Python out from the does the thing so Nelson as my developers I didn't some so actually we need fixes is that it walks but it doesn't look good no pretty region which is better and uh this granted regular Piaget so that's opened up also and you have here the Opel the show up in the up on 1 of ocean deployed right now so so if we can create a pull request so we can get my some change going so we can then the in general can do that 1 so are in the tub lingo Apple requested risking my attempts to integrate my changes into of the line and so I I did this proposal delivery came up in and took over in automatically is starting to walk on it to make sure is for as you see we have banded doing our work force which is apropos so here we go will go ahead and open over here and see where the statuses this are multiple stages inside delivering a walk through this you don't know on but the 1st stages of F of the verify approval stage this is what human looks actually at the code you can go and a common inside it's kind inside the PRB you don't actually have to use the PR to do it as a background bi-directional communication but I'm just going to go and say you know what I just know someone pretty damn this thing to do OK so we're learning from Angel intestinal syntax that's checks to make sure that no should works right that's probably important so pale cool right on this event the good thing to know is going to run prevent steps relationship but if we show you what we have here go so will go back to the P here just keep watching here you know going through the bill phase built those does it again deliveries normally designed for multiple people in multiple teams working on 1 1 system so you have multiple people pushing the odds and at the same time to make sure that's the manager of the later stages to so now also hear his us to work on things and we're going to go through the Lincoln went and a syntax check 1 more time to make sure that nothing strange underneath so it's Pastora to master well before we doing this yes you can i pointing and we're doing it again and Morgan image we had to acceptance of this is what's gonna do for us and acceptance is is going to build out a pound before moving the code over if you look here as you see all travels being served search served at this 1 location a moment when we get to do acceptance will actually spin up see another version pop up without no coated with no traffic going to select a get visually inspected to see that's what we want to be and we should increase addition circuit and then we also want to earn my division aspect but also use automation and because of this traffic is not his real alive to the customers but it's you know where is we can actually apply all or automation into that for us for no private luxury and make sure is actually correct in all the 4 that passes them will migrate to the traffic in for this simple then we're going to this might be the traffic all at once but you can do all sorts of scenarios that can make you migrate part of the traffic in the Canary testing as you get me 3 per cent of the traffic to the new system see how behaves is up to you to define but using the I. Piaget and the API calls or in any way you want our how these dispositional right here for the sake of time that we would to flip everything was he's all goods in a single shot energy Shearer per-share at the acceptance rate now and which real the see here we also get the little chart updated here started having to another 1 in the engine was developed from what is it serves to mountains just functional acceptance June was 1 of the cell so we talked about how you do so this is really the world models and a lot of power the the the ability that you might have you know about where the law happened it is always the the fact that the delivery of it would be a lot things all the way useful and to which brings the trend on the word recall that from what J. and the call out here I thought grew engineers were good and they're always get things right the 1st time the way of what I was right so each shirt what are we talking about here that the and so that's a happening as you see it actually failed and that's were looking for this example and what we'll do is ask Nelson an incident effects of 4 so something goes wrong and where it was
not so there will have to make is that this moment head in the video that the only if
I see that OK so so it seems I actually but at the wrong about the link over there which is good because our official application is it still sitting right so know about our time and work with it and then it is making either the that that is easy to that is our production site that a formal pop platform 1 whatever and so we haven't actually heard our customers which is also right so we actually in the process of pushing aneurysm code we didn't ourselves in the foot yeah so so as I push another up-to-date another attempt to the power the the approximate and if I had had the leverage that already taken anyone is action run that and make again make sure that everything is good Our in we wouldn't be electable mother tries so they're both shoulders is broken but now we want to try again in a few passes then you actually ruled out to production but 1 of the things that are especially like about to delivery pipeline is the law may develop a toolkit what's so I should I wanted this coder direction trust instead through OK I'll prove the effect that if you look at work OK there's still time never mind
what somebody else's fault it's CDC and in the because of the fact that you absolutely and we will get their own skin of the and pick up the good trying to be entertaining giggles right now well what way we can go back there where the this 1 is doing things so you can go back then showed the new new new and improved look and if it's go up I had to in my in my task is here I projection the wrong place I was as you pointed to a wrong Europe but my my validation and the rules that I created that checks if my was that might vitamins good say is not then there's delivery to say stop this guy before he does anything what's right so that's what happened here although I was good to use it it death but with all the links with them and so of the security check that that if you have noticed we do actually have but we do have some security where we did not run because the basics of abroad of the Samaritan which of them are how securities little the but she never know what effect so you we can go on them but if you go to wizard have the so the basic
idea out in this process you can actually keep all these history and when we deliver all this history get section together it is also came back directly as leader for the virtual like OK how did I got to the sport like what kind of changes he came to rise in this case here our we the ectoderm my fix that was that I was pointing to the wrong place no hope would be is the right of with it the parents to published the book was facing this 1 specific 1 is research will spin up another instance inside another versions of AP and to myself the among we like grain OK doing acceptance again to make sure that it's all this that the validation reward that to be is that you can you know in a way that for proved nominal DOA just get there the the all the the new year few I think that it is to break it the here again functional this is the minus is not happy isn't it it's this slope here like what it's a during the next 2 the we conclude that it Starting from the yeah so the way we do to these them appears at without we created a cookbook that the has actually been called the unit test all this but think can be secured In using another cookbook that that we wrote about the Google wrote the true religion really is that the abstracts AP as a as a chef resource so I can now say but this act I define my work full time in chef In just say that that this is the right and and there's no Nelson delivered grand check by the 1st delivery was kind of so this time to go ahead and deliver through don't have because of particle current and as you can see here we go and then the pick it up in the 2nd and finished and skid not to the union of union phase if you don't know is a phase that make sure that so everyone's approved up to and I view the code looks cool union is a is a dirty environment where all the all the people working on the code come together so make sure that actually works and it works I expected to the next phase after union which is called a rehearsal is designed to be another phase where you would induce something like another instance where you blow all this stuff away and rebuild it from scratch locally because reusing App Engine we don't have to worry about that we can just more or less kind of scoot through to make sure is still open passes or are our basic integration testing of basic unit test the view the advantage of a passive of different product and so on scenario as you can see we we don't know the point this nature it up there and our running through a friend from functional to which no you yes and no 1 reversal which vectors another phase on which is designed for ground up but this in this case because reason past we can is kind of be like a nose just another union dementia should doesn't break the to to do the although sometimes those phases they they they look alike elements of I think and 1 of the important part for them is even if and I to merge walks the mean the final chord is actually correct right so I can't limit that's my my my my my breast like works perfectly the process and then you test or passive but in when you bring them together they fail so that's why there is this extra step to make sure that are not when you bring everything together they still have that there would be 1st the Library of and prize I each of the what the on this section of the universe holier understand all the world to the unity the sort I think it is this point so the title of the the other 1 was and this think of that but it's still inside you and I the they in many cases also that you will have a bit of a single what the agent application have enough Peter that sickle database and maybe a container the running you but it is equal to what you would use a sequel if I do know your sleep a but at that's a good time for you to that's right so when you want to make sure that different from and we're almost done I think we are so low that we we she her everything rain and they had no we think that history of the the sum of the the OK and what's refresh
led a that's the today show so you have to do that with the i I agree this is much much better get used to room 1st in leverage this should have is
that right so once again I have alive demos for these guys to there's always going to be a all little hooks in that so I guess it and a number of things you wanna throw a g here in this in this session because once again we've done a lot over last year but we only have 1 session so I think arms and so lot you guys and girls had said you've used GCP may morass a follow-up question why photoproduction workloads running on GCP that prices said economics in that but and is also part of the reason when they go through a few elements of GCP had to tell you why you should start running production workloads with us and I'll keep this short and production time and return a reference from cascade talk about the multi colleges that might be something as of interest as well for sure but I
think the 1 thing we were trying to make sure we message the folks I Brennan said this morning and I thought it was unfortunately a very strong statement that we don't really do marketing very well sometimes but when it comes the stuff we build these amazing products we have this incredible cloud whatever have but we don't really tell anybody about this expect you to find it and tried out I think it's awesome so I think they're doing events like this in getting in front of folks and sharing this with you know hopefully will change that will really try to do is instead of getting you think about Google like this we want you starting in like us
it's not obviously just search phrase there's
actually 7 different Google products they each have over a billion active users of peace so you think about the global infrastructure footprint actually power all of that is pretty significant right is a Diane Greene on SEP and you just started time at the end of last year was the original founders of the EM where she says something very simple but I think was varied poignant if you
will a lot people say that we've only been doing this for 2 or 3 years but that's actually fundamentally false OK we've actually doing this for like 15 years we haven't sold at it is a very important thing is I read a lot of blogs 1 analyst want folks on twitterverse and what have you and all this kind of Google to 3 years to 3 years there's a difference between selling it and like making money on that and actually being operationally sound and delivering we just decided a few years ago star making money and book clock we assign less externalize what were doing for 15 years and doesn't make money but is still fundamentally different argument right and so I think some of the things
we always like to point out is our global network OK it's our fiber thousands of miles of fiber wrapping the planet if we were a telco would be the world's 2nd largest telco so over
100 points of presence all over the globe here's why that matters to you folks OK if you have requests coming from your applications and it hits 1 other POPs that Google was inside and Google sees is destined for you will resources is such that request and are backbone rises the holy to the destination in and the whole way back to the user that's a lot less harsh is running on our fiber the results of the performance this example where infrastructure actually does matter right what's
so in terms the platform itself but lots of you the log and so on were used you know we have you know most 30 services now a services Virtual machines on your Isaac everybody else I will point out our data services which is 1 area where we are significantly differentiated because once again is stuff that Google's been doing for 15 years in the original MapReduce white paper that became the do emission big table before k it goes on and on on DRAM and so on so forth these are all services the were now externalizing for people to consume so data Proc have and me I don't want that the labels on here because if it's not fair to this expect you know all the products doesn't matter the data Proc is basically host and to do you spent 11 thousand cures in 90 seconds when you're jobs blown away you pay by the minute for what you use idea when you knew that missing block goals you don't count the the other who have a does the the work at would do more a day but big query picked bakeries are petabyte-scale data warehouse of you think like I am your traditionally UW security actually when we're customer has 40 petabytes of data this is inside of the query ingest 2 million rows 2nd into our cloud-based data warehouse the did multi-terabyte queries In literally 10 seconds completely unheard of but nobody really knows were going on a platform right so I invite you have nothing else announced knock it off with the cell Sketchpad if nothing else the data services will possess a before usually logic Valderrama do great things Release 3 4 or 5 so again this is definitely area where we we think we have a little bit of of a fun road Fourier nav Causey application runtime services and things like when Nelson J. so 1 thing I do I'll point
out is as 1st compute engine goes because I figure for this audience of fighting the most interesting us are virtual machines right reems arms good times for folks at tried a platform of the pay when you try it out In the last 6 months we've got our Abu times down to us and they ask what would you think a faster timers for AViiON beaverish shown number 2 frame number 10 minutes the this you specter 2nd can't move the expectation that him very fair and suggest that at so doubleness in any manner death who gets a minute when find 5 minutes mes-a-mes OpenStack OK enough value is that it it's so we're down to 30 seconds now we were 45 seconds about a year ago were down to 30 seconds us very long story is 1 attribute the cascade is well that's a wonderful story were 1 of our engineers were an executive briefing sold of executives right use up talking about our machines and he's like yeah you were kind of our version machines motion because we got down to 30 seconds we were trying to get better you there around they're like a case that normal or not normal so once again it's something when you think about the elastic scale or a spinning things up tearing things down these the things that matter we assert machine configurations Ringius 680 thousand I ops that's not a typo 680 thousand read 360 thousand right and there are discussed in the EMS selecting your exact CPU uh CPU and memory actually so teacher sizes we have that to you can pick from preconfigured machines but yet she drag a slider pick the exact CPU exact memory and will tell you over time will your performance characteristics are of those the Gene go back and right size the and as you got we see customer saving 15 20 % doing in because they don't have to waste CPU or memory based off of what they had to pick so lie migration another big 1 and this is pretty simple is it a but here the nose the I'm on the sphere and as I DRS and the motion it's essentially what it is when we take hardware offline you don't even know about that this literally completely obfuscated from anything happening infrastructural it's a very cool use
discounts I about pricing I sometimes don't like talking about pricing is once again on the blogosphere and with analysts and what not they make it sound like users trying to be cheaper OK but what we're actually doing things like sustained use discounts so sticker price yeah wouldn't be cheaper but was sustained use discounts duo is the more use machine we discuss more because then we can plan for what our underlying infrastructure looks like so says once you get over 100 psych 20 or 30 % a month we search slap in more money off the price it happens automatically you have to sign a deal you have to sign contracts to get this or anything it just automatically happens you can shave up to 53 per cent off of that virtual machine Madison anything and well I don't really run on IBM's 24 7 365 we have she added for you once again automatically we going inferred instances so this the same machine tive in your spilled up and down up and down another 1 this the same size up and down up and down waxy IRI demersal give you the next step 2 pretty cool stuff there's just marketing so storage and were
cheaper and so I'm back from the storage perspective so are object storage so we we do other served like everybody else reprisal the cheaper but the 1 area that we really once again and I want to let you guys know what differentiation is with archival storage thing long-term like slow-moving blocks of ice like that kind of long term so or tape were tape for T I think that the thank you for that and so we got near line OK so concerning as that a penny a day a month pricing naturally answers but here's a key it's the same exact API footprints so few coal if you code to our standard uh Storage API and then you wanna code to ah Archival Storage API it's exactly the same use change with the tears it's 1 property that's it you're done worse even a bigger deal it is is time to 1st bite you don't issue an API commanding come back 6 hours later after lunch and and 3 5 seconds the 1st bite to even talk to customers they're using this because the price point for some of their standard web year assets because maybe it's it's it's all use case driven right maybe it's not noisy user to have to wait 3 to 5 seconds and the a set of save a boatload of money on the back and storage seeking to creative things like that as well and implies
security others had this really quickly because I am surprised I shouldn't be I was surprised and how many folks is defaulted to Google the rainbow colored logo and doing consumer and that not taking security seriously here's what I will tell you if we engage with you and if you reach up to google we start having conversations we bring in the heavy guns and I dollars and I promise you that our story around security when you think of the assets so we have to protect and natural were extending to what you're consuming off of our platform security tends not be an issue we were 500 engineers nest just engineers the focus on security 24 7 do we found hardly before anybody else even knew about we were patch before the public announcement but it is the kinds of things we do day in and day out so there's a whole day half a day a full day we can spend on this by just wander through the nite out there because sometimes for whatever reason people default to we will discuss this fun little thing hanging out there security is not an issue so sorry that was
a little too much not went too far in the high level marketing the stuff but I wanna make sure that everybody understood that if you bring about Google Cloud if you logged in and tried it out yet defies choose this is something very real frugal is something that for us we were running all of Google on society were shut it down like greater as you know the 1 in the essay again but but look at least I threw it out there and good thank you that it that's OK 5 halves of have ideas we what powder go from we don't do that to hurt you know this was a keynote this morning or rosiglitazone program so I show you ample time for a friend from cascade might was gonna come up and tell multicolored story about what they've been doing customers and health will be relevant to some of things as having a plant cell next up you have a lavalier of this yeah they get quite a bit of flak for
that 1 by the way so at some point it out I'm gonna go through this this fairly quick
because I like to live demo of a real world scenario I which emprical
stuff going running so I Michael nor the Cascadia for those of you that's the donor
Cascadia EOI would basically In a demobs cloud engineering services company what is that mean well it basically means that if you're gonna do something on the cloud you but do it if you're going to actually on me right we firmly believe that it's all about the automation so we often go into organizations and help them actually understand how these things like chaff ahah get things deployed get them up and running in the cloud of and we also demand services so after the fact which means we people engineers in bed you guys can help you continue building out will have you back so you guys get getting good sleep is proposed to it but it i'd so the example I'm going to go through the set is kind of a real customer scenario that the 1st it is what a call out is we're a little bit about uh the CAO and at Alaska talk about how he had his business outcomes and I will say the number 1 thing we see when people these being successful is they focus on the endgame of what a by really to learn the business not it is really cool technology by now the world's acknowledges in a room so we all love the technology wanna tell the story but when you then try to go get them to give you money to do it and they give commitment to it you always have to tell different stories and then I can give you a demo that very quickly takes you through and the story in itself so this kind of represents you gotta have a vision of the need of overall we try use a company how you got to have a business architecture we try and achieve over Rory arching technology architecture and an employee who were about the technology in self so I'm going to try to tell that story kind of
responsivity with with what most of us won't necessarily consider that exciting which is non profits now the reason why this 1 is very interesting is we're not talking about 8 1 single error nonprofit we're talking about the G 8 Summit but basically mandated that all nonprofits open up a much older research sir collaborating all that well as you imagine you have all these different organizations that are basically storing information everywhere and then nowhere communicators mattered most information is sitting or things like laptops a lot of it is sitting on on pieces of paper so it becomes very difficult on but really from a vision perspective that that's how do we actually use technology star sharing all of this and action do for global good and that's really the story that that I'm going to tell here so what is realistically the
problem the problem is that there is an literally petabytes verses of and worth of data to scattered across all these different kinds of organizations everything from when also the collecting it is everything from little you know in at that the devices sitting scatter across with sensors on the you pulled in geospatial information that is in the petabytes of and also what I'm actually the demo is suppose I'm standing somewhere in Africa and there's actually no Internet rather than using paper I'm gonna use a little telephone device and basically be able to on the fly submit information and then have that available immediately for the whole world the see as well that will we're gonna demo in this specific case I don't sharable published the scope of of the people come in and how can they immediately published and start looking at it with various different analytical tools and how do you provide that completely integrated way in such a way that their business so the people there a researcher can on the fly actually deployed these applications and without even calling i t pretty powerful stuff because underneath the hood is all the tool when we talk about today and the reason this thing is moltaí cloud is it every cloud provider has some specific thing that makes them extremely valuable but how do you tie all these things together and 1 of the see the secret answers that is actually chef is a real key component and for a lot of this work so what I'm going to demonstrate and this looks really messy but it can be very hard to uh kind a shows the public or clouds and see its axis storing a lot of information that's coming from what multiple different organizations and it's not the were controlling the plane but we are taking a look at all information indexing and pulling together on this and curious ones in there like joy and so because to do a lot of geospatial Miguel de la calculations that should do in line right there along with it who is a very key component some other guys there too I'm including using things like private storage allow that's because less information other so kinds of rules around does it have to stay in a country data sovereignty and all that stuff has to be accounted for as part of the ODP as you can see I have there's different tools they'll plug up to where I'm actually the demo what's called the OTK Open Data kit and and that S 3 a big analytical tool on the automation bit is 1 also demonstrate in terms of which chapters to this kind of an oversimplified picture but I try to make a simple occurred and it actually the confusing so all this that I it said this is the scenario that I'm actually go through and which is basically the 1st thing I'm going to do it is the researcher's gonna log in and they'd want to set up that field research and they want the filler should be in the run around them this device right here on but they also want completely pre-integrated aware were about how amateur makes us into some kind of storage mechanism that it's backed up and it's safe and I don't mean of to worry about how that actually is connecting the other and so that's where the through here so the 1st thing I'm going to do if the
Internet works correctly it is accidental log and this is the the simplified portal I should say so we actually had the run this on a kind of separate instance on but it's a simplified portal and as you can see I will only have a couple things sourced portals search and the 1st thing I'm going to do it is just to make sure ICM a type in demo because that's actually the research were going to do in the field and it doesn't come up with anything just to show that have any tricks up my sleeve if to this search portal this is where I would go to say I now need to have that research in the field going on and the business would go in here and the base of it is deployed the new data source and they give it the name of some sort this pick something like the city is here now I actually now choose what application 1 member of a tire emodi case a lot of no OTK unaware that is and I wanted you that this is gonna go selected and then selects what cloud provider actor provides us in this case and actually OTK runs prenatally and cool it actually is 1 of the the really nice things how well the integrated with it I get to pick what cell which is extremely important and the reason why that's important is if I'm standing in Africa I don't want the latency of dealing with the server in US LOD get as close as I possibly in the that was gonna pick has pick US Central 1 I can also not pre-populate so not only can it sustain server I can move in the utk I connection also pre-populated with forms as being created in the past so that means that not only am I getting application but actually getting functionality already know that has the surveys and that I want on its and like dental and I also get the select final is the last 1 how Alexa when run and run in development is case but think of it this way if I'm picking production is can have very different rules in terms of moving all this data and start making it available to a broader audience times the pick development for simplicity and I hit the floor in cross your fingers nothing went wrong he start spinning so right now what it's doing it's it's actually spinning up a server if you don't believe me it's sitting right here
in so it's actually now gone out and it's starting up a whole new server and it starts to go through everything you need to do to configure that serve now it's gonna take i don't know not 5 seconds probably a good 30 seconds for the will bed and a probably just a little bit more for shaft actually but basically configure the server itself
so I just wanna know what's going on right now and it's going to chicken scribble little bit but what's going on is that the automation is kicked in the Google is has been set up and the server that the these recall Chef Server shift service figuring out what cookbooks are coming in and it's now star that configuration 1st is going put in OTK make sure that set up appropriately make sure this completely configured is gonna go pull those forms which is an decay specific thing that it's doing that would published against out and setting the whole thing up and it's actually gonna come and tell us when it's done so right now we all have
the said wait now what we're doing that we can place of elevator music that really and book I just have to configure this thing right now and what going to could figurative about the
object storage you'll see that I to get a ura l that I can now call for that server but if I'm a on the but on the business side so I just have to type in the year after giving a 2nd Ch but is with fingers the and that it's the good time at that to unite in the hacker conference that should be effectual of of they II so yeah you get little attack that 1 and so as you can see here that the CPU 1 is active going out this because a to a lot of processing right now and you'll see a dead when it when it finally gets in the place but we have to wait years 2nd Tennessee this well I will actually warned that right now if I go in actually look at the instance does to your points and then we look at the the logs i exiting is completed its run from the group perspective which so now it's just students were now waiting on ship that is 1 point that out who was gonna stay out that that you can have the effect that the Zappalà this is where you know you just can't have the said wait a little bit but it's actually extremely quick if you if you think about going in setting of entire server filling out the application pre configuring everything and actually making it so they can communicate into this entire network of a moltaí Cloud solution by and then includes words sector going to store this information so the interesting thing here is that it's actually we using Google for the application but it's actually storing in AWS and the 2 actually interconnected complete so it's seamless that we have and were about anything and everything is done on the fly side we set the institution together manually everything is automated right in the more all right here
so the was and
this the you know what the the
lot all of the and so on this is what you the the but would be that so the you know what the the that the the 1st is the use of the please the so all that CCI actually now launched beauty K is up and running as no data and every 7 felony forms I've actually just filled in the forms on here on the device Amanda me standing there next some big line filling in a form right 1 and then argues that hit save and exit and it's already been shipped off and so if all of you in here is I refresh this is now what I'm having to do here real quick if we're gonna take it a step further is a match it is going to generic the system so it decays set so that it up loads to it every 8 minutes and we're really not waiting in minutes so I'm going to go on a forms lists and this can have basically force it to release the data and need update a couple things and it can publish and make sure that I'm not some random individual doing something and I can now see that ever get to publish this data now here's where it's important if I go back to this member this whole you either the working with on the business side I cannot go in and I can actually search for that form narratives my so this is basically where the capturing the raw data so everything on capture this device is now available stored in in the cloud in this case AWS and is now available for consumption and was going to make sure is that they're not yet takes a 2nd chance there it is sure that so here's what's important is now I can take that data and a matter somebody standing there on the ground the entering information and immediately somebody with let's a sitting in the states and then say OK let me actually look at what's going on the
ground and I'm using S 3 in this case a Methodist but I'm using a plugin for S 3
as you see here comes the point brisket time and the Pritzker lovers when I say
they and because it actually also had spatial information it actually pinpoint me where I'm standing right now right but imagine all the information you can capture and the business does not know anything about what's going behind going on behind the scenes right and as the talk about Alaska Airlines is it should be completely invisible everybody but what's going on behind the scenes is a tremendous amount work in automation that to seamlessly brings together so I'm that's basically it so I had hopefully that makes sense the if
Streuungsdiagramm
Subtraktion
Bit
Drei
Systemplattform
Binärcode
Term
Streuungsdiagramm
Offene Menge
Bit
Benutzerfreundlichkeit
Güte der Anpassung
Systemplattform
Quellcode
Cloud Computing
Raum-Zeit
MIDI <Musikelektronik>
Open Source
Dienst <Informatik>
Minimum
Endogene Variable
Debugging
Tabelle <Informatik>
Parametersystem
Datenmissbrauch
Sichtenkonzept
Datennetz
Graph
Güte der Anpassung
Versionsverwaltung
Plug in
Sprachsynthese
Cloud Computing
Systemplattform
Service provider
Mapping <Computergraphik>
Virtuelle Maschine
Kollaboration <Informatik>
Differential
Rechter Winkel
Mereologie
Strategisches Spiel
Wort <Informatik>
Dualitätstheorie
Verschiebungsoperator
App <Programm>
Wald <Graphentheorie>
Kartesische Koordinaten
Physikalisches System
Softwareentwickler
Telekommunikation
Momentenproblem
Mathematisierung
Versionsverwaltung
Zellularer Automat
Kartesische Koordinaten
Inzidenzalgebra
Gesetz <Physik>
Division
Code
Videokonferenz
Multiplikation
Informationsmodellierung
Datenmanagement
Softwareentwickler
Bildgebendes Verfahren
Phasenumwandlung
Schreib-Lese-Kopf
Leistung <Physik>
Soundverarbeitung
Softwaretest
Lineares Funktional
Addition
Güte der Anpassung
Programmverifikation
Einfache Genauigkeit
Systemaufruf
Physikalisches System
Bitrate
Ereignishorizont
Quick-Sort
Teilmenge
Energiedichte
Twitter <Softwareplattform>
Forcing
Rechter Winkel
Digitaltechnik
Mereologie
Wort <Informatik>
URL
Repository <Informatik>
Soundverarbeitung
Web Site
Prozess <Physik>
Computersicherheit
Gruppenoperation
Validität
Formale Grammatik
Schlussregel
Kartesische Koordinaten
Biprodukt
Binder <Informatik>
Systemplattform
Gesetz <Physik>
Code
Richtung
Task
Projektive Ebene
Message-Passing
Leistung <Physik>
Bit
Subtraktion
Prozess <Physik>
Gewichtete Summe
Punkt
Komponententest
Natürliche Zahl
Mathematisierung
Versionsverwaltung
Fortsetzung <Mathematik>
Kartesische Koordinaten
Element <Mathematik>
Code
Eins
Dämpfung
Reverse Engineering
Programmbibliothek
Vererbungshierarchie
Inverser Limes
Grundraum
Phasenumwandlung
Nominalskaliertes Merkmal
Softwaretest
App <Programm>
Sichtenkonzept
Chord <Kommunikationsprotokoll>
Abstraktionsebene
Datenhaltung
Validität
Biprodukt
Quick-Sort
Integral
Teilmenge
Arithmetisches Mittel
Rechter Winkel
Beweistheorie
Mereologie
Garbentheorie
Programmierumgebung
Message-Passing
Instantiierung
Beanspruchung
Befehl <Informatik>
Demo <Programm>
Multiplikation
Mereologie
Zahlenbereich
Element <Mathematik>
Biprodukt
Cloud Computing
Ereignishorizont
Message-Passing
Parametersystem
Subtraktion
Maßstab
Web log
Rechter Winkel
Virtuelle Maschine
Cloud Computing
Biprodukt
Leistung <Physik>
Resultante
Punkt
Datennetz
Datennetz
Kartesische Koordinaten
Urbild <Mathematik>
Punkt
Rahmenproblem
Data-Warehouse-Konzept
Versionsverwaltung
Zahlenbereich
Zellularer Automat
Kartesische Koordinaten
Zentraleinheit
Zählen
Systemplattform
Term
Mathematische Logik
Virtuelle Maschine
Erwartungswert
Prozess <Informatik>
Migration <Informatik>
Delisches Problem
Konfigurationsraum
Zentrische Streckung
Hardware
Computersicherheit
Güte der Anpassung
Zwei
Rechenzeit
Abfrage
p-Block
Migration <Informatik>
Biprodukt
Cloud Computing
Fourier-Entwicklung
Dienst <Informatik>
Navigieren
Flächeninhalt
Rechter Winkel
Festspeicher
Charakteristisches Polynom
Tabelle <Informatik>
Magnetbandlaufwerk
Streuungsdiagramm
Virtuelle Maschine
Kondition <Mathematik>
Speicher <Informatik>
Instantiierung
Design by Contract
Magnetbandlaufwerk
Umsetzung <Informatik>
Punkt
Mathematisierung
Term
Systemplattform
Code
Benutzerbeteiligung
Differential
Perspektive
Speicher <Informatik>
Default
Gerade
Streuungsdiagramm
Dean-Zahl
Kategorie <Mathematik>
Computersicherheit
Zwei
Magnetbandlaufwerk
Speicher <Informatik>
p-Block
Fokalpunkt
Packprogramm
Objekt <Kategorie>
Patch <Software>
Flächeninhalt
Standardabweichung
Bit
Zellularer Automat
Optimierung
Übergang
Streuungsdiagramm
Bit
Demo <Programm>
Punkt
Zahlenbereich
Systemaufruf
Dienst <Informatik>
Cloud Computing
Hinterlegungsverfahren <Kryptologie>
Dienst <Informatik>
Menge
Reelle Zahl
Unordnung
Computerarchitektur
Maschinelles Sehen
Ebene
Telekommunikation
Bit
Demo <Programm>
Subtraktion
Selbst organisierendes System
Gruppenoperation
Kartesische Koordinaten
Term
Internetworking
Eins
Multiplikation
Perspektive
Notebook-Computer
Zusammenhängender Graph
Speicher <Informatik>
Maschinelles Sehen
Gerade
Kraftfahrzeugmechatroniker
Streuung
Schlussregel
Rechnen
Cloud Computing
Software Development Kit
Arithmetisches Mittel
Datenfeld
Offene Menge
Rechter Winkel
Automatische Indexierung
Mereologie
Information
Einfach zusammenhängender Raum
Trennungsaxiom
Lineares Funktional
Bit
Subtraktion
Demo <Programm>
Zwei
Zellularer Automat
Kartesische Koordinaten
Schlussregel
Quellcode
Biprodukt
Cloud Computing
Term
Quick-Sort
Internetworking
Bildschirmmaske
Datenfeld
Portal <Internet>
Datentyp
Server
Softwareentwickler
Instantiierung
Umwandlungsenthalpie
Dienst <Informatik>
Bildschirmmaske
Bit
Rechter Winkel
Gruppe <Mathematik>
Server
Konfigurationsraum
Verschiebungsoperator
Soundverarbeitung
Bit
Punkt
Datennetz
Gruppenkeim
t-Test
Kartesische Koordinaten
Zentraleinheit
Login
Objekt <Kategorie>
Menge
Perspektive
Datentyp
Server
Wort <Informatik>
Information
Speicher <Informatik>
Hacker
Ganze Funktion
Streuungsdiagramm
Instantiierung
Motion Capturing
System F
Bildschirmmaske
Matching <Graphentheorie>
Last
Mailing-Liste
Information
Physikalisches System
Cloud Computing
Gerade
Aggregatzustand
Demoszene <Programmierung>
Punkt
Plug in
Information

Metadaten

Formale Metadaten

Titel Chef Journey on Google Cloud - July 12, 2016
Serientitel ChefConf 2016
Autor Asghar, J. J.
Jung, Brandon
Araujo, Nelson
Norring, Michael
Marshall, Jay
Lizenz CC-Namensnennung - Weitergabe unter gleichen Bedingungen 3.0 Unported:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen und nicht-kommerziellen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen und das Werk bzw. diesen Inhalt auch in veränderter Form nur unter den Bedingungen dieser Lizenz weitergeben.
DOI 10.5446/34624
Herausgeber Confreaks, LLC
Erscheinungsjahr 2016
Sprache Englisch

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract Learn how customers are using chef and chef delivery to run, deploy, and manage workloads across Google Cloud on PaaS, IaaS and container environments.

Ähnliche Filme

Loading...
Feedback