AWS reInvent 2021 Unpacked

okay should we uh should we get going norm yeah sounds good so hello and welcome to our webinar with cloud academy on aws re invent 2021 unpacked uh first i'd like to introduce everybody on this call my name is noam bunder i'm a content architect we are joined by stuart scott who is our aws content lead we also have will meadows a senior content creator with aws we have jorge negron aws content architect and uh one want to let you all know that this is being recorded so if you don't have a chance to watch it right now we will send out a link to the recording we're also taking questions in the chat box and we'll address any questions that you might have at the end of our talk so uh let me kick this off myself i'll say this is my first reinvent that i've attended and uh i was really just impressed with the amount of people that showed up you know coming out of covid knowing that everybody was vaccinated it was just nice to see so many people there a lot of great energy a lot of great excitement especially in all the keynotes and just a huge huge amount of attendees both the venetian palazzo the win was really great so um i think i'll direct my my first uh question over to stu stu uh stu what what did you think about some of the new uh compute and storage services that were announced yeah sure thank you now um just to follow up from what what you said about the the sheer number of people i remember my first real event it was just it was a little bit overwhelming i wasn't sure what to kind of expect but you're walking through the corridors of these hotels and it's just jam-packed and these are wide spaces um it's just people everywhere all the time so but yeah i'm glad you're glad you had a good time um i didn't make it this year but hopefully next year i'll be there in person with the rest of the crowd academy team um so yeah back back to re invent and the services and the new features etc there was a ton of stuff announced across loads of different domains for analytics machine learning networking application integration and a ton more um i just want to touch on some of the storage ones um that you may have seen announced during the keynotes or just kind of soft releases in the background so i just want to run through some of those storage ones the first one i'll talk about is the new offline tape migration using um aws snowball edge so this is essentially a new capability that sits within the snow family and part of the snow family you have the snow cone the snowball and the snowmobile and these are really used to help you remove the constraints such as limited bandwidth and high connection costs or high latency when you're trying to transfer data into and out of aws so this new um tape migration with the snowball edge you can use this to get your physical tape in into aws so you would order the the snowball it would arrive and then you can migrate your tape infrastructure onto snowball and then send it into aws for them to load on the cloud um once your your tape backups are in the cloud in aws probably s3 um then you can use the virtual tape library that integrates with aw storage gateway to then access that tape data and these these snowball devices they handle up to 80 terabytes of storage um it's secure it's rugged and it's trackable just like the rest of the the snow family if you're familiar with it at all and it's really easy to use using the aws management console it's got a very easy ui um especially if you've not used the snowball family before um but yeah it's just a great way of getting your your tape data or your old physical tape into the cloud without relying on an internet connection or kind of restrictions on network bandwidth uh capabilities and and costs as well um so that was the the first one um that kind of caught my eye the next one was to do with amazon s3 and obvious backup so aws backup now supports amazon s3 for those unfamiliar with backup it's used as a centralized service to to help you manage your backups across different resources and automate your backups across different services and it uses policies to to govern what should be backed up and and when um those order using aws backup will know that it already supports services like ec2 dynamo db tables efs ebs neptune and there's a ton of others too i think there's like 12 or 13 services that integrates with now but the announcement was that it now supports s3 so you can backup your s3 buckets using aws backup so you might have application data that you want to have in a centralized backup which has been stored on s3 and you can use aws backup to do that and you can create different backup options as well uh for example you can create a point in time backup of your s3 bucket or schedule backups over a given time period and it's very easy to restore as well as a couple of clicks and you have your restore of your your s2 backup so it that's what aws has done like this reinvent i think they've simplified a lot of things a lot of features and a lot of services so you'll see things kind of getting easier and easier which is kind of good but also you want to know what's happening underneath all of that which i which i kind of like were any of the um were any of the services sort of renamed uh yes i'll come to that in just a moment actually when i uh so i haven't haven't finished with this there's a couple of announcements on it so we've got some renames of of storage classes um but just before we move on to that um with regards to the backup aws backup you can use existing backup plans that you already have um that you already use and you just add the s3 components to it and it also integrates with aws organizations as well so you can use one backup policy uh for your sd backups across multiple different aws accounts that you have so not leaving s3 we've got some new storage classes and name changes as well um in the world as f3 s3 so the first new class is s3 glacier instant retrieval storage class when i started announcing these new storage classes straight away i thought people are just going to get even more confused as to what storage class to use because there's already a list of about seven or eight and it can be confusing sometimes to know which storage class you should use for one particular workload so um just to let you know going forward i'll be creating a course on this to help you really understand how to manage your workloads on s3 and which storage class you should use so keep a lookout for that um but this s3 glacier instant retrieval storage class is basically used for use cases where you need long term storage at low cost but you also require that uh immediate millisecond retrieval of that data should you need it um and this sits between two other storage classes it kind of sits between the um infrequent access of s3 and then the other gracious storage classes now the next one that was announced was a part of intelligent tiering it's a new intelligent tier and crafts and this is the archive instant access tier so this tier has been designed to be used for workloads uh either unknown or have changing access patterns and it's optimized for data that isn't accessed for kind of months at a time rather than days and automatically optimizes your storage costs for early access data that needs instant retrieval now for those that have used um intelligent tiering before you will know that there's two other um tiers in there we have the frequent and the infrequent tiers so this is the third tier for the intelligent theory now what happens is if you have an object that's not been accessed for 30 consecutive days then that object is moved from the frequent tier to the infrequent access to and this is a cheaper tier so it's trying to save you money now what we have now is that when an object hasn't been accessed for 90 consecutive days then the s3 intelligent team will move the object from the infrequent access to the archive instant access tier with a cost saving of approximately 68 i say approximately that's pretty precise for that's what they're saying it gives about 68 saving um now the last element um is a name change of the existing storage class of amazon s3 glacier storage class which is now called the s3 glacier flexible retrieval so you can see that there's quite a bit of confusion here i think will happen for people so it's worth taking a bit of time to look into this now this existing storage class although it's been renamed it does offer some additional benefits it gives free bulk retrieval um in 5 to 12 hours and the storage price has also been reduced by 10 in all regions and this would be effective or was effective from uh the first of december so there's three classes of glacier now you have the s3 glacier instant retrieval which is for long-lived data that's really accessed but the data is available in milliseconds so this is kind of great for news media you have s3 glacier flexible retrieval backup uh so it's um sorry that's great for backup vr and large amounts of long-term rarely access data but retrieval costs um aren't an issue for you then this is great and if you don't need the data back straight away then that's also good because it's not instant retrieval and then you also have the deep archive which as you probably know this is the the lowest cost storage um in aws for archive storage um and you can restore data in at least 12 hours so that's kind of the s3 storage classes uh moving on from there it's just a couple more points on storage that i just want to touch relating to ebs so there's two um kind of smaller feature announcements of ebs which i think are pretty cool uh you have the recycle bin for ebs snapshots now so this is a great feature um for those of us uh that might accidentally delete ebs snapshots that we should have kept which can happen you can get a lot of ebs snapshots there and you think oh do i need it or not and you kind of delete it i need to realize that yeah you did actually need it so you can create different retention rules to retain deleted snapshots so that you can recover from them um like i say in an accidental deletion uh you simply set up some retention period in those rules um against different ebs snapshots and you can also base this on resource tags of those snapshots as well to capture a bunch in one go and it's a good idea as well to use different iam users to manage the retention and restoration of ebs snapshots because then this generally allows you um to create a two-stage deletion process so one team might delete the ebs snapshot by accident but then you have another team that looks after the restoration and management of the recycle bin of ebs so you should always be able to get it back and then we also have amazon ebs snapshots archive and this is basically a new storage tier for ebs and your snapshots it allows you to keep certain snapshots for months at a time or even longer uh you might want to do this perhaps for compliance reasons um however this storage tier allows you to do it more cost effectively storing your snapshots in this new archive um class can help you achieve up to 75 percent lower storage costs um but the restore for your archive snapshot will take between 24 to 72 hours so that's just a roundup of some of the storage elements that um that were announced like i say some of them kind of soft release in the background not in there not in the keynotes and others were so i just wanted to give you a quick rundown of those um um the s3 intelligent uh tiering uh how does that sort of compare to just using generic life cycle policies this might be a question someone would ask in general um so so your lifecycle policies will uh move your objects from one storage class to another but the intelligent tiering is one storage class itself so you can you can move it to intelligent tiering but then the intelligent tiering element takes over and then manages that those objects for you based on when those objects were accessed so probably good if you don't you want a more hands-off approach i would assume yeah definitely definitely if you're if you're unsure of the kind of the access patterns of that data um and you want to try and manage your cost a bit more cost effectively then you can put it into the intelligent tiering and then maybe over time you can review those objects when they were and weren't accessed and then move that data to a more appropriate um storage class to save even more money and so this new addition to the intelligent hearing is just another level to save money so i don't i can keep my hands off of it in general that's me yeah that's awesome yeah yeah that's a really great announcement about that that uh the the option to move the tape backups off off off site i remember always being impressed by these big robots that moved the tapes around you know back in the day so yeah it's just a great way to get your your tapes into aws and then like i say you can then leverage the the virtual tape library with the storage gateway to then access that data easily and and retrieve it and restore it whenever you need to i did have um the preview of the stuff i was looking at which was machine learning stuff they actually lumped in some some database and this is kind of storagey so i want to throw it in real quick if you guys don't mind yeah okay so dynamodb got a new storage class as well which was standard and frequent access table classes so this is very similar to s3 in frequent access where if you've got some data that you need to hold on to but you don't want to poke it that much uh this new way to store your data with the standard and frequent access class for dynamodb is up to 60 cheaper um so for example a gigabyte of data in a dynamodb table would cost you 60 cents this new stuff uh would be 25 cents uh per gigabyte so that's not bad at all um oh sorry it's 25 cents for the standard and 10 cents for infrequent i have a different number for this other one the part where you get charged more obviously is for your rights and reads so if we were looking at the on-demand uh cost for a right it'd be um 1.2 dollars per million rights and then um 1.25 or 0.25 dollars per million reads as opposed to this is where we get charged back with the new table class one point uh five six and one point point three one respectively so it's about a thirty percent more money on your reads and rights but you save sixty percent on the storage itself so something to think about uh and that ballpark yep yeah so so uh jorge what uh what are some of the services that you saw in terms of announcements from keynotes or perhaps i see a i notice a a number of services that are a lot about increasing the reach of aws or aws know-how to to a variety of individuals so the first observation that i noticed was very significant is the addition of a new pillar to the well-architected framework this was introduced by amazon cto for aws verner vogels this is the sustainability uh pillar and this is impactful because the well architected framework document is a reference document that gets quoted repeatedly in aws certifications exams and it joins us the six pillar where you have operational excellence for example security i usually speak about security first and foremost but you have operational excellent security reliability performance efficiency and cost optimization this is basically a gathering of knowledge that aws puts together in the form of a number of questions that you can ask in order to perform the discovery process to know what situation is it that you're in in terms of your current it implementation also questions that relate to best practices the way aws has accumulated through experience uh by working with different customers so this sustainability pillar is a way to permit now customers to join in this environmental best practices for computing that has been in the news the greening of i.t there's a number of phrases for it i thought this was a very innovative it's simply a way for aws to share and reach a broader set of individuals so along those same lines of expanding the reachability i notice also aws amplify studio now this is just a tool integrated tool the benefit here is that it permits full stack development within aws with the amplify suite of services and in this case you get to you get a visual development environment to manufacture the user interface uh with prefabricated uh elements of user interface and you can import edit as needed but more importantly it allows you to wire everything from the ui all the way to the back end using the amplify suite of tools so this is a way for developers perhaps back-end developers now to be a little more involved with the front-end become full-stack developers that's an expansion and skill set and expansion and opportunities a way to perhaps create a better quality result simply because you're able to maintain cohesion between the different layers usually that entails communication between different teams so i saw that one as a kind of democratizing another area of reaching more individuals is the introduction of the new software development kits and we're talking new sdks for swift which is apple's preferred language for software development these days it replaces objective c as a was considered now a legacy language hasn't received any modernization since the late 80s maybe early 90s so it's been around for a while swift now it's replacing it as an efficient syntax easy to learn is being used as the first programming language for a number of elementary school children kotlin consider it java java's grandchild in that it's efficient in syntax is java compliant it can import java library so kotlin is also a language that is modern it's efficient it's compatible with java's gaining a lot of momentum and last but not least rust rust is actually the type of language that is very memory efficient uh it's efficient in terms of the footprint that it takes uh doesn't have any garbage collection memory management is highly highly uh optimized so rust is getting a lot of popularity even even the dr vernon vogels paid some attention to the rust sdk so there are a number of services that have come up that simply make things easier for developers uh easier to develop the front ui another example is the aws software developers kit it's now on version two and and and the big the big takeaway there is that if you have used cdk version one you most likely got used to have to import uh the packages for each of the services that you wanted to use you know to create a i guess a smaller footprint uh in in the final executable now everything of that has been bundled into one import and the idea there is that if you have a complex application that in terms of the cdk is tapping into anything more than a dozen services the whole import packaging becomes a little bit of a heavy burden so now one single package will allow you to tap into any service that you want and that's an efficiency mechanism right then and there saving time saving effort and then last but not least construct hub think of it as a repository for a prefabricated functionality that is compatible with the cloud developers kit it's compatible even with terraform developers kit a number of other cdks that are available so construct hub becomes a repository a registry of prefabricated functionality developed using this cloud development frameworks so those are the services that i notice kind of take a little bit of a step in the right direction as far as being able to include more individuals individuals that perhaps didn't have any background expertise individuals that are now preparing certain tools like swift kotlin languages now are able to get a 100 percent into the aws ecosystem fantastic thanks jorge and uh will how about on the aiml side have you seen any uh interesting announcements that you'd like to highlight yeah there was some neat stuff so uh i think i'll open up with talking about uh sagemaker ground truth plus let me just say there was like a lot of additions to sagemaker on uh on this update not a whole lot of new services in the machine learning realm but definitely sagemaker got a nice boost the boost we enjoy for machine learning i think in like previous three events sagemancer always gets um some attention i think it was yeah it was the same this year as well yeah it's it's the child you know it's like all right come on our our chosen one let's see if we can what was that what was the nice part okay so a little bit ago we had a ground truth come out for sage maker and this is something that can help you with labeling your data sets you've got images that are coming in from your cameras or from wherever and they've got dogs and they've got people and they've got cars so that stuff doesn't come pre-labeled so either you painfully by hand do it yourself or you have your team of interns do it or you use ground truth and ground truth will take your images and send it over to amazon mechanical turk which is a someone else's different paid team of interns not yours and they're cheaper and then they will label the data for you the problem is if you have very specific uh data labeling like i don't know medical imagery uh specifically checking for maybe bone cancers or fractures or something that was specific that you wouldn't want to trust to some random person on the internet to label for you they've got this new ground truth plus and this new update promises to have a team of experts we'll see how well that works um go through your uh your labeling and and try to do it well for for you over mechanical turk so the way this works uh in order to do this you have to fill out a form aws will give you a call they will set up um a meeting with you to discuss what it is that you're trying to get labeled on your training data you give them a little bit of time they'll go find the experts for you and eventually you put all your training data into a bucket um and then at some point it starts and then once it starts the experts will go through and start to do the labeling that you would expect as normal and then a machine learning algorithm is following them and watching all the things they're labeling as well and so as they go the algorithm itself and the model will start talking ahead of wherever they are in their task and try to label it for them and if it's correct obviously the experts will give a thumbs up if it's not they'll give the thumbs down and relabel it themselves and so theoretically though it may cost a little bit more up front once this little thing this this new model has been trained by the experts it should be able to do the rest of the job for them or at least speed up the efficiency that's kind of cool um theoretically i i think this will work really well this is one of those things i'll have to see how it works in practice so i don't know i think that's neat don't think it's neat it's fantastic absolutely that's one of the biggest parts when it comes to be able to classify material through machine learning the whole concept of labeling what is exactly this thing and a lot of times it used to require human intervention now automation is just continues to improve make it easier faster better yeah i'm curious to see how much it really cost because i couldn't find anything specific on the page so maybe it's a job by job that they'll give you a bill for it but oh neat um what was another cool one i want to talk about let's see ah yes sage maker canvas so this was a neat one that let me give you the premise so you've got somebody who's like a business development person or a business admin who's got like a sheet of let's say product data and your products are like every item you sell uh where your customers were that you shipped it to whether it was on time or if it was late what the distance was that customer that you shipped it so you've got this excel spreadsheet right you can shove it into sagemaker canvas which is an entire drag and drop machine learning interface so they say and it looked like it um where you can plug that in and make a quick little model i can do predictive analysis on that for you without having to have any machine learning background without having to have any real aws experience so they say um and it can start making predictions for you so you would plug in another data set of maybe new uh possible orders and it could give you a prediction where those would arrive on time or not uh that was the example they gave and just having the power to have this paint by numbers machine learning for i don't know introductory machine learning for business people i thought it was very powerful to use the amazon speak it was democratizing machine learning for more people which is always good it's it's not super friendly to get into from the face of it so that was nifty i recommend taking a look at that one yeah i think i think that's pretty good i think that will be well received within the industry as well because more and more things are moving to ml but sometimes you need that specialist knowledge to kind of get started and go in but they're making it a lot easier a lot easier yeah that's a well speaking of that to make it a little bit even more easier is they release sagemaker studio lab which is um so without any aws account without any credit cards and with zero knowledge whatsoever aws will give you this studio lab account for free there's probably some like qualifications in there where they will give you 8 or 12 hours of cpu time and 4 hours of gpu time per session to enjoy as you please to just play around in the machine learning space this is basically setting up a jupiter notebook for you or like that where you could download from github any of the uh already pre-built uh things from there and you can play around and mess with it for free so there's no worrying about damaging your aws uh credit card bill uh with whoever you're playing around with or your account your boss's account the company's account so this is nice like if you just want to go home and uh sign up for this and they say it's free so you get your 12 hours of cpu or gpo time to to play with the all the frameworks like pytorch tensorflow and all the other good stuff so if you don't know anything about machine learning go watch a quick youtube video and then go try to get a free account and go play with it for free that's what i recommend that also seemed really cool and i guess my final oh go ahead good no i said i like that one absolutely it makes it easier i can add if you have any questions there's actually a new service that was introduced aws repost and this is basically a replacement to aws forums and it aggregates the collected intelligence of aws customers partners employees folks like us uh that have been back and forth in cloud and aws on a regular basis so it's basically a service that puts you in touch is even plugged in directly to aws support and the idea here is that if you have a question you get a legit answer you get an answer that is useful an answer that is applicable basically improving the fidelity as opposed to the noise when it comes to these services i wonder if it becomes something like a stack overflow you know the reference for developers to go and ask questions and get logical intelligent usable answers so aws repos is about to prove itself and it was introduced this week in re invent that's awesome i like useful and applicable those are those are valuable uh i guess the last thing sort of finish up my spiel is aws is throwing out some scholarship program uh funding for the underrepresented and under uh served high school and college students so if you are one of those people on the call i highly recommend you go check this out it's free education free money uh i'm i guess i'm the free guy today so i like uh i know i like machine learning so if i can get the free and the machine learning together we've done well so i recommend this one go take a look at it it's uh if it works for you give it a go and uh i had a question for stu uh about the uh some of the new compute features that you've seen uh announced yeah there's um there was a bunch of stuff announced uh in the compute side of things um for new processors to to new instance types so one of the new instance types was the the ec2 m1 mac instances which are used if you're looking to build uh and test applications for um ipads your macs your apple watch your phone basically anything apple um apple tv as well so they used to give you um a great option to quickly provision mac os environments in aws now the mac instances were announced last year i think and i think it was a last last room event but this year they've basically expanded their instant types to support mac os and ios workloads with this new m1 mac instance um and it basically lets you access machines built around the apple designed m1 system um system on chip and they deliver up to 60 better performance um against the x86 based ec2 mac instances um from a hardware perspective um they're powered by a mac mini which has the m1 chip with eight cpu cores eight gpu cores 16 giga memory and a 16 core apple mule engine in addition to this they are also based on the aws nitro system uh with 10 gigabits of vpc network bandwidth and 8 gigabits per second of ebs storage bandwidth using a high speed thunderbolt connection so that's a new instance type on on the mac side uh we've also got a couple of new storage optimized instances as well we've got the im4gn and the is4 gen these are both powered by the graviton 2 processor and the key characteristics for these instances is that they deliver high disk io performance with a ton of storage capacity as well we've already got a number of different storage optimized instances uh such as the d2 the i2 and i3s uh the d3s etc so this is just another one that has been added to the instance family i've got 30 terabytes of nvme storage using aws nitro ssd devices and this helps to reduce your your i o latency by up to 60 percent um are used to support high speed access to large amounts of data now the im4 gn this is ideal for applications that demand large amounts of dense ssd storage and high compute performance but they're not necessarily that memory intensive uh the is4 gen um instances have been built for apps that do large amounts of random i o to large amounts of ssd storage um from a performance perspective they're both more cost effective and better when compared to the previous i3 storage optimized instances we've also got another new instance type the ec2 g5g instance which is also powered by aws graviton 2 and there's a lot of stuff a lot of ec2 instance families got a lot of storage classes yeah it's it can take a while to get your head around them i don't think anyone's going to really know every instance type but anyway going back to the g5gs um so they extend that the graviton 2 price performance benefits to gpu-based workloads including graphics applications and machine learning inference etc so some of the use cases can be streaming android gaming graphics rendering autonomous vehicle simulations etc and they also feature from a hardware perspective uh an nvidia t4g tensor core gpu uh with 25 gigabits per second of networking bandwidth and 19 gigabits of ebs bandwidth and these are available in six different instant sizes uh starting from the xl through to 2xl4 816 all the way up to metal being the highest performing of them all and then i just want to highlight the ec2 c7g instances and these are powered by the new graviton three processor and that was announced so the graviton 3 is new it's a custom designed by aws and it's 25 faster than the existing graviton 2 processors and it also achieves better price performance than the previous graviton processors as well and they have twice the amount of floating point performance when running compute intensive workloads so this is kind of a great choice if you're looking to support and run high compute workloads um for example ml that will's been talking about um today uh batch processing scientific modeling etc and like i say this is a new instance that now runs on a graviton three processor uh i think it's the first one um so yeah great choice for your hpc stuff your gaming distributed analysis um and it's the first also instance type offered by aws that comes with ddr5 memory and this gives you enhanced speeds um for accessing any data stored in memory uh compared to the c6 jit the c6g instances it has a 50 increase in memory bandwidth and also has an increase of network bandwidth of 20 running at 30 gigabits per second and it supports elastic fabric adapters as well so i know that was quite a lot of information and not a long amount of time but i just wanted to highlight that there has been a number of new instances out there so we have the new instance it's powered by the new graviton three processor which is faster than the graviton twos we've got some instances that are focused on storage optimized and also um the gpu instances as well oh as well as the mac ones that i started with which i've already forgotten about so yeah there's quite a few there it's quite a few there um it's hard to keep track of them because and they thing to remember which was in one of the keynotes all of this is customer driven these these instance types are basically requested by customers workloads that they are implementing so uh basically we heard that a lot yeah we had that last of the keynotes and i think it was in uh werner's um i think he said when we got like 200 services it's basically all your fault um implying that it's all the requests from the customers that you know you're asking for this this is why i have so many uh yeah yeah that's a really good thing it really impressed me is just the wide range of industries and and use cases and and business use cases that are driving the proliferation of these services and the services conversely are allowing these all these different kinds of industries and business use cases to let to leverage the aws uh uh compute power and storage power like like you mentioned earlier so really impressive yeah so so we have a few questions from the audience uh i'll take uh here um uh here's one one question here although i did not attend this year it seems to me that new announcements did not have the huge wow factor of the past reinvents it feels like aws has reached a level that focuses more on refinement so what what's your take so uh anybody want to take a stab at that yeah i kind of agree with that and we was kind of discussing this offline before um this webinar that i think there's been a lot more enhancements to existing services rather than the big dramatic you know new service launches i mean that has been a couple obviously for example the aws private 5g um which allows you to scale a private mobile low latency high bandwidth network in a matter of days um and this simply allows you to implement your own cellular network um for your resources so that was a pretty big announcement and i think some of the stage sagemaker ones that uh will was talking about the sagemaker canvas i think um it was um so that has been some definitely new services out there which are pretty cool and and also we had a bunch of analytic ones um lose serverless uh i think it's redshift serverless rush of serverless uh emr serverless and i think there's msk serverless as well so there has been some some really cool new services out there and but i i did feel like it kind of focused on more enhancing um services yeah there's a big breakthrough in the in the networking space with the announcement of the private wan feature i thought that was a pretty that was a big wow in fact oh yeah we also got the cell phone service that was kind of neat absolutely i was going to point out this year the department and in this particular case the the aws private 5g uh is is completely new and breakthrough this is this is going to be disruptive if you worked in the telecom industry then you know that 4g lte is competing with 5g sometimes even confused there's a little bit of a tug of war happening with regard to spectrum uh the etas the latency for you to be able to provision the correct equipment with the right vendor to obtain spectrum licenses to be able to connect it all together and integrate with your existing environments it used to be a proposition measured in months possibly years of latency for you to be able to go from let's do this to okay it's finally complete so this is one of those few innovatives the aws 5g brings 5g network type latencies which are load latencies high bandwidth and it's basically already pre-provisioned with spectrum it's already pre-provisioned with the equipment your console you define what is it that you want where do you need it how many units are you going to be deploying and then basically aws ships you the equipment you insert the sim chips uh on the gear and light it up and use it made it pretty easy disruptive to say the least anyone who has done this prior to this service uh becoming available can tell you it's taken a very long time and this is going to democratize the 5g consider you're doing a conference and you need high-speed low-latency wi-fi this is the type of service that can make it happen in a matter of days or weeks as opposed to months or years yeah i thought also the whole amplify studio and and just the whole ecosystem of amplify provides a level of accessibility that that is sort of unprecedented with the recent proliferation of more and more you know javascript licenses that javascript sdks and libraries that you have to kind of get involved with both on the front end and back end so it's almost like an easier barrier of entry to both front-end and back-end developers that we haven't seen in a number of years and that's that was a pretty big wow factor for me as well absolutely i saw also a pretty good introduction which is the aws mainframe modernization and i know that sounds a little bit of legacy but what's new here is that that's yes the type of work that is happening with frames these days number one is suffering from a shortage of uh specialists most of the folks that were involved in these implementations are now within retirement age if they're not already retired so there's a shortage of competent personnel to address mainframe workloads and uh the mainframe modernization platform again is just an accumulated set of knowledge and know-how that aws is making available along with the development tools for you to port your mainframe application to let's call it a mainframe emulator for the sake of uh of a better term you can move it to aws or you can choose to refactor the application the golden nugget there is a recommendations engine that gathers the expertise aws has accumulated with a number of agencies including the new york times and it gives you a become a recommended plan it analyzes your what you have uh it tells you give you a little bit of forensics what's going on and it gives you a recommendation if you want to port do a lift and shift you may want to approach it this way if you want to refactor you may want to approach it in this other way and it provides your compiler to translate from cobol or pl1 into java which is pretty significant so that's a little bit of old technology that has been revamped with a new service very innovative and it's going to basically open the accessibility doors to new talent to now pick up on these legacy mainframe workloads let's see if we have any other questions here um so uh what was the announcement that impressed you the most for me i think it was pretty obvious it was canvas i think this is a great spot to get into if you're if you have to run numbers for people and you have to make predictions and you're kind of curious about machine learning and the wizardry behind the box i think this is a great place to get into it and i highly recommend it i know i'll go give it a test here once i get some downtime from writing uh to go play with it yeah i think that yeah yeah just on that note well i just want to kind of uh bring up the point of um as a team we're gonna take all these kind of features and announcements and review over the next couple of weeks or so and just put forward uh a bit of a road map of new content that we're going to create other courses or blogs um labs etc uh just to let you know that we're going to be focusing on all these new releases and getting some great content out there for you to help you kind of travel through the forest of all of these announcements and and help you on your way to start using these services and features as well in your own production environment so just keep a look out for that going forward we're going to be writing new courses and writing blogs et cetera and using these first hand so um we can kind of uncover the little bugs and know hows and how to's for you to make it easy so yeah yeah and well you know along the line the same lines of you know getting getting these types of uh hands-on experiences cloud academy provides a tremendous amount of labs besides the content that stu mentioned and really serves as a skills intelligence platform for you to be able to sort of build up a skills database and do analytics on different gaps you have in your organization in terms of education and training and then sort of being able to tactfully attack that and develop learning paths for for people who are using the platform so we encourage you to check out cloud academy and and see the different types of training we have available let's see what other questions we have um so what what new service do you see yourself going to implement first so anybody here uh see any particular service that excites you the most i i quite like aws security and all the different security services that there are so i want to check out that didn't you yeah i might have done yet um so yeah i want to check out the the new um inspector service because that's had a a big revamp so i'm going to have a little bit of a play with that and and hopefully get some content out for that fairly soon very good so uh another question we have is uh the new elk stack uh any quote on this so elastic uh okay so uh uh another another question i need to look yeah very good uh what new apis and services are are you most excited about that's very similar to the the previous question but like in general i think anything that is helping it helping people get into the cloud easier because the cloud is the future and so is machine learning uh i think is is great so any of the new tech that came out that is just a slight ramp into this this space i think is great uh and really should be like praised sounds great okay wrapping up here i think we might have one or two more questions here uh so aws exams uh have a six six-month delay for new features so in cloud academy do you follow this uh strictly this rule for pre-courses yeah we tried to stick to this rule so as uh new features and services are announced we try to get the content out there as quick as possible and then we have the ability to just to integrate these lectures and courses with the satisfaction learning paths when they start finding their way into the questions and as a team we we regularly sit the exams just so we're familiar with with the content in there to make sure that we are generating the the learning paths that have all the information that you need to set the exams so yeah that that's definitely something we we follow i agree with stuart another absolutely we stay on top more importantly for any reason you take a a a module you see a module and you notice that something is dated or inaccurate you can always report it support cloudacademy.com that's going to arrive to one of us and it's definitely going to be paid attention to and we're going to mitigate it as quickly as possible uh in the next revision for the course so you also have an input if you see something that is you can always contribute and help but we do stay on top and we try to for example the second version of the solutions uh the sysops associate the moment that that exam was released in july we had a learning path already prepared for that particular exam that now has a hands-on component so that's just one example of how that has happened previously and has happened a number of times also with the content in all platforms and uh if if you're already a cloud academy customer you'll probably regularly see when you're completing a learning path that you'll go back to it and it'll you'll see a message saying some content has been updated et cetera et cetera et cetera and that's where we add in either new courses new labs uh update existing content um to make sure that it aligns with the blueprint of the exam guide here's a question for you sue around the topic of security so do you think that uh some uh some of what aws is doing is competing with security vendors more aggressively especially with siem vendors i think there's always been kind of a partnership between um aws and a lot of the vendors um you do kind of see year-on-year new features and announcements coming out in the security space i didn't think there was as many this year um i was kind of hoping for a few more new security services if i'm honest um but it seems like they kind of dulled that down yes this year obviously there was certain security features across different services um but from a new service perspective i think that was only really the new inspector and i think there are some advancements made in shield as well um so i i think they they kind of tend to partner with a lot of security vendors and there's a whole list of approved security vendor partners as well so i i don't i don't think massively at this stage how about you will and jorge would you agree with it absolutely i will have to go also for this reinvent in particular uh security focus for the most part on enhanced network uh internet of things type security and improve security for containers one of the more significant is what stewart mentioned the introduction of the new amazon inspector the the older amazon inspector is now called inspector classic which is aws way of saying the time will come where this uh will be deprecated in fact the classic network is um coming to an end soon um i believe yeah if anything i think uh a lot more um options to uh have been exposed in the security field that you know in terms of apis data exchange that you could sort of tie in uh from from a more a larger variety of vendors very good so i think we're coming close to uh wrapping up here uh uh any any other uh comments here from from from jorge will stew um yeah yeah i think we've covered a lot of things if you have any questions on um any of the announcements that we've spoken about today and feel free to get in touch uh if you want to see any new content of any announcements or features that you've um heard through the previous week then also please let us know and we can kind of hopefully build that into our roadmap going forward but yeah it's been great chatting with with you guys uh kind of discussing reinvents announcements there's a whole ton of um features and elements we we didn't have time to discuss but would be here all day i think if we if we went through everything so but hopefully it's helped some of you guys uh for joining the call uh on the call so thank you for very much for making the time yep very thank you for attending this uh this uh webinar from cloud academy and uh please look for the link of the recording that will be shared and if you have any further questions and uh please uh come to cloudacademy.com and we'll be happy to answer them for you thank you have a good one yes thank you bye bye

Channel: Cloud Academy

Views: 312

Keywords: #cloudacademy, AWS, Amazon Web Services, AWSreInvent, AWSreInvent2021

Id: 4IhuiAJ3lt0

Length: 52min 27sec (3147 seconds)

Published: Mon Dec 06 2021