[MUSIC] Everybody welcome to another
exciting episode of data exposed. I'm your host Scott Klein and
back with me today is Louis, Louis how are you? >> Hi, I'm good thank you,
good to see you. >> For those that didn't
shamelessly watch the last show why don't you take a second
to introduce yourself. >> My name is Louis I'm
a principal programmer manager on a SQL Server team. And I work with Azure to make sure
it's the best platform to run SQL Server in the world. >> Woo, so this is the guy
we love over here, yeah. So you were here probably
about six months ago. >> Something like that, yeah. >> To give us kind of an intro and
to SQL in an Azure VM. And we were talking before, let's talk about kinda what's
new since then, and then kinda, from the portal experience and the
kinda things you've been working on. >> Totally, so we have been doing
a lot of good work on Azure to make sure that SQL is very,
very easy to use. And we're helping automate and
manage a bunch of things for SQL Server. So yeah, let's talk about that, what I'm thinking to do is let's
go through the portal experience. Show some of the new things that
we have added and maybe we'll talk also of some things we're adding
in the next two to three months. So we have the Azure portal,
as you show here. By the way, I have a couple of
virtual machines here running already SQL,
probably just show them at the end. My lab, this is test labs, for
people who happen to use it, great way of managing lab
environments and auto-start and auto-shut down your machines
at the end of the day. So it's going to just create
a new SQL Server machine. As you show, I could but
I'm gonna go to the long paths. I'm gonna go into Databases and
then in Databases, I'm gonna go and see all the database
offerings that we have. And so here inside of the servers
we have SQL Server, right? So we're going to click on SQL
Server and here we have the list of all the images that already
come with SQL pre-installed. Now, the interesting thing here
is we have all the different versions and additions that
we support for SQL Server. So customers who are using today
SQL Server 2012, or even 2008 R2, they can still go and
have a virtual machine running with those versions of SQL Server very,
very quickly. >> Can we take a quick tangent? Cuz I noticed the icon, or the yeah, the tile below that says
SQL Server AlwaysOn Cluster. So what you have highlighted now
is just the specific SQL Server that spins up a single VM, right? But I think that's pretty, how
recent is that one AlwaysOn Cluster? >> That's probably like five or
six months ago, probably soon after
the last time I came here. I can show that as well,
and we can talk about that. That basically will configure
a full OS deployment end to end. >> Yep,
if we have time we'll do that, otherwise maybe we have you come
back and do that, because I think that would be an interesting
to go through that end to end. >> Totally, for people who want
to have a high availability and search and recovery,
that's a good way to do it. So we have as I was saying all the
different versions of SQL Server, we have different editions as well. So you can see here in this screen,
for example, we have Enterprise Edition,
we have Web, Standard, etc. We can go all the way to SQL 2016,
right, so the latest and greatest of SQL Server. And here we have again Enterprise,
Standard, Web, something new that we introduced
is a couple of images. So now we have SQL Express. So for people who are not
familiar with that. This is a free license SQL Server. You can use this even for
your production workers if you want. You don't have to pay for
SQL licensing, but it's for light workloads really. >> This is for very,
very small workloads, all the way to 10
gigabytes of storage. It's very,
very small one gigabyte of memory. So if you have a website who doesn't
have heavy traffic, you can go and run SQL Server for
free with this so great to have. The other one that we have is
SQL Server Developer Edition. And this has everything
that SQL Server offers. >> Yeah, this is because
of like Service Pack One, it did that programmable service
area across all editions. >> Exactly. So just to recap, we released
probably three or four months ago, SQL Server 2016 Service Pack One. And for the first time we started
adding a lot of the features that we only had in SQL Enterprise Edition,
also in Standard Edition. So, for example, things like memory
tables and column store indexes. And some of the security features
that we have in SQL Enterprise are now available in
SQL Standard as well. Having said that,
there's always higher limits in terms of the performance
that you can do with SQL Enterprise. So the number of cores, for
example, goes beyond 16 cores and the amount of memory,
it is unlimited. So whatever amount of memory you
have in the virtual machine, you can use it. And there is some as well. So some optimization in terms
of query processing and things like that. SQL Enterprise,
still the best edition to use. For Enterprise workloads however if
you have some smaller workloads, you can still benefit from
a low dose Enterprise features. >> Yeah and it's interesting
because we have a lot of ISVs and other customers that go hey, you know they're running
Standard Edition, right? Yes, Enterprise you wanna, if you
have your Enterprise workload, we have a lot of people go,
hey are these features in Standard? So I think this is awesome
that we support that. >> Yeah cuz now the priority
surface is the same. [CROSSTALK] Go on and change and get better performance
than with enterprise. But you don't have
to change your app. So for people who are writing
a new application for the first time or they want to
test the new features in SQL 2016, the best way to
SQL Developer Edition. Again this is another
free license image. So I don't know if everybody is
aware of this but you can just go and create this virtual machine
with SQL Server Developer. And then you don't have to pay for
the SQL licensing so SQL licensing is free. You can go and do as much development and
as much testing as you want. And then at some point if you
wanna go and move into production, then is where you will have
to look into SQL Standard for example or even SQL Enterprise
depending on what you want. >> That basically look really good. >> Definitely a great
place to start, so those are definitely some new ones,
new images that we have. The other one, a big one that we
have here is now SQL on Red Hat. So as you know now, SQL Server
is coming to Linux as well, sometime this year. And as part of that,
we started making it easy for customers to go and try SQL on
Red Hat for the first time. So we now have an image that
comes with SQL pre-installed, running on Red Hat Enterprise. >> Do we have telemetry data that
says how many of those are being spun up? Are people looking at that? >> Yeah, no, there's a lot,
a lot of interest. So I just pointed this, probably
thousands of virtual machines, they're running on SQL in Linux
at this point in time on Azure. There's a lot of interest,
especially because still this image doesn't charge for
SQL obviously because it doesn't. And doesn't charge for Red Hat
either because we have an agreement going on with Red Hat right now
where they don't charge because this is still for both SQL and-
>> Very nice. So yeah, at some point later this
year, we're gonna GA the offering. We're gonna have the final
version of SQL Server. Another point obviously, we're going
to charge depending on the edition, Enterprise or the Standard Web,
etc, but right now it's all free so people go on and try SQL on Linux. The other final thing that
we wanted to show is this, you see this BYOL thing? All of them have started with BYOL. You may wonder what they are,
well BYOL stands for bring your own license. And these are images that come
with SQL Server installed as well. It could be
SQL Standard SQL Enterprise, but they're not gonna charge you for
the SQL licensing cost. You can still use these for production, but what this allows
you to do is bring your licenses. So for customers who today have
an on premises licenses, right, so they already have some licenses for SQL Server running in their
corporate environment, they can easily bring those
licenses into Azure and run that SQL Server on
Azure using those licenses. And a little quickly,
I'll show you how this works. The only constraint is this bring
your own license benefit is for our software assurance customers. So customers who have this software
assurance benefit also called the Enterprise agreement
with Microsoft. They have access to them and
it's very, very simple. So all you do is click on them and
all you have to do is within ten days you need to
notify Microsoft and then there is a way to
do it electronically. And you have to notify that you
are gonna bring a certain number of course or license course
of SQL Server into Azure. And that's it, I mean, they
basically come with SQL install. Obviously you need to have
an interface agreement. >> [LAUGH]
Does the compute price change if I'm
bringing my own license or? >> No,
the compute price is still the same. So the cost of running the virtual
machine itself doesn't change, and that depends on the size
of the virtual machine. So obviously if you have
a machine that has 32 cores and half a terabyte of RAM, it's gonna
be significantly more expensive. Then you have like one or two
core machines with maybe three or four gigabytes. >> Yeah, I was just wondering since
I'm bringing my own license but that's okay. >> Yeah,
the licensing part is for SQL. Having said that, you can do
the same with Windows Server. So people who are running
Windows Server today on premises and may have licenses through
software assurance, they can bring those
as well too to Azure. >> Compute prices just
across the board, right. >> Compute prices across the board,
that doesn't change. But yeah, just recapping, so we have
SQL 2016 including SP1, office. We have old edition SQL Server,
we have developers. >> Yeah.
>> People who want to start development of testing. And then for we have Express,
we talk about SQL Domain Notes, and obviously the images. So let's go and
create a virtual machine with SQL. Let's go and
try SQL 16 SP One Enterprise. It's a new deployment model that
we have which allows you to manage groups of resources together and
do things like resource control, access control and things like that. Let's go and
create a virtual machine. Some of these I'm
gonna go quickly and some of these things
we're very familiar with. I'll set my SQL server virtual
machine and It was a number. The disk type, so at this point
we'll default all of the SQL Server virtual machines to a disk. This is because we want to
offer the best performance when you try
the SQL Server experience. Great, less than four
millisecond latency. If you are running straight
from the local cache, we have local cache with
these virtual machines. And if your database fits
inside of the local cache, Is less than one millisecond later. >> Wow.
>> So super, super fast. Having said that, if all you want to
do is just test the basic experience and you just want to functionally
see how it works, you don't care too much about performance at this
point, and you want the lowest cost, you can always-
>> Yep, just playing with it maybe and
just experimenting with it. >> You can always go and change into
traditional hard drives or drives, definitively not as fast as this. You can have always functional test. You can try that. >> Okay, very cool. >> I'm gonna specify my
administrator user name. Some password to make sure
that I wrote it twice. Yes, got it right. My subscription, so this is where
I want to create my resources. And there is. This is the logical grouping for
all of my resources. >> Yeah.
>> They can start them together or restart them, pause them,
delete them, whatever I wanna do. I'm gonna say my SQL group and
the number there, and the location. And as you know, we have more locations in
the world than anybody else. >> Mm-hm, very cool.
>> So you can choose Australia, Brazil, Canada, Asia, Japan, Europe. A bunch of places in Europe and obviously a bunch of places
in the US et cetera. We are gonna select US Because
it's the closest one to where we are right now. And then we're gonna move into
the next part which is choosing the size for the virtual machine. >> Yeah.
>> This is where you are gonna pay for the compute cost. And you specify the number of cores
that you want to run from SQL server to memory. How many disks. These disks today can be all
the way to one terabyte. We are actually gonna
increase the size. >> Wow.
>> You can store more per disk, but for right now it's one terabyte. So if you see for example this
machine DS14, it comes with 16 cores and then total useable memory
32 terabytes of storage. That's a lot of storage. 50,000 IOPS definitely good for
must applications. It has a local cache of 224GB,
right. >> Okay.
>> This means, if part of your database or ideally
most of your database fits in that. That means that all your
reads are going to come from cache of the machine. >> Okay.
>> And again, just one millisecond
latency is very, very fast. Obviously if you use things
like compression, for example in SQL, or color mystery
indexes, that's going to compress significantly the size of both
the tables and the indexes. >> Yeah.
>> And that means that you're
probably gonna fit four or five times more of your database
into that cache system. >> Interesting. Yeah that's cool. >> Have even better performance. >> Okay. >> For people who want bigger or
as says, we have a little more. So you select. And you have a lot
more options here. People who need, big, big,
big machines, we have things like the GS five for example
with 32 cores, 448 GB of memory. >> [LAUGH]
>> It's a little more than- >> Go big or go home. [LAUGH]
>> Yeah, service. 64 TB of space
[INAUDIBLE] scenarios. You have a local [INAUDIBLE] of-
>> [CROSSTALK]. And this is actually, I forgot to
mention, this is a local [INAUDIBLE] you can use for whatever you want
like just a temporary drive. This is [INAUDIBLE].
Like you can put 10 TB in there for example. >> Yeah, okay. >> [INAUDIBLE] working
on [INAUDIBLE] here, the size of the local root cache.. >> Wow, okay. >> So imagine with compression or column store says you probably
can fit 16 terabytes or so. And that's why, I don't know
if we mentioned it before, probably not because it was recent. >> [LAUGH]
>> We announced I think October, November, that now we certify this
machie, GS five with FastTrack. FastTrack is a data warehouse
certification that the secret server team provides for
different hardware vendors. We just certified this machine. This machine is certified for
16 Terabytes of data warehousing. >> Wow, okay. >> Basically this means the SQL
server team says if you're on data warehousing in this machine
up to 16 terabytes. There's actually some kinds you can
go all the way to 20 terabytes, but we recommend 16. We promise you good performance. Anyway, you have a bunch of machines
there for whatever you wanna use. For now let's stick to this one. We're gonna use 14 machine,
again, good for most production work that's
out there, 16 cores. 112 gigabytes of memory. And then we have some general
settings for the virtual machine. Probably you don't need to go and
change these things. If you want, you can change things
like do you wanna create a subnet, for example, for your network and
things like that. You can create things like a public
DNS name so that you can use a name to access your machine over
the world anywhere that you want. >> Wow, okay. >> You can really set
a public DNS name. Then some diagnostics, it's always
good to have by default for example to understand
what is happening during. The [INAUDIBLE] diagnostics,
it's good to enable this, cuz it will allow you to machine. >> Yeah, okay. >> And using for Windows and
SQL Server as well. So for example per second. Things like per SQL, page like. So we're gonna say okay. And then we get into the SQL server
aspect of the virtual machine. >> Okay.
>> Some of these we've touched last time, but some of them are new, so we're gonna go through
the whole thing quickly. First thing, SQL connectivity. How do you wanna connect to SQL
server in this virtual machine. You can specify local, which means
you have to remote to the machines or RDP to the machine and
it's a local. So it's like the high
level of security, probably you're gonna get used
to some basic testing and you don't wanna open the machine for
anything else than USRDP. Private, that's probably the most
commonly used setting and this is, you allow connectivity
only within a virtual network. The virtual network is a battery, a logical, secure battery for
your network where all the resources inside
the network can see each other. >> Okay. >> All right, so
you have example that website and you want the website to use SQL
Server, than you put the website in SQL Server [CROSSTALK]
The same secure bannering. Also people who have VPN tunnels,
for example, you can configure a VPN tunnel
through your physical router. You can have something like Cisco or you can do it through even software
for example, Windows Server. But you can establish a VPN tunnel. And notice the other communication
is secure between you on-premises environment. And you'll be. >> Okay.
>> And the same thing will apply so you select private
connectivity to SQL server only machines that are part of your
corporate enviroment will be able to connect as far as. >> Okay.
>> And then the final option is public. And this offers connectivity
over the internet. And this is for people doing the test for example
that is what I'm gonna use here. If I wanna connect from my laptop to
Management Studio and I wanna go and do some development
over the internet, and I want to connect from everywhere
in the world I just specify that. >> Okay.
>> Obviously I can still specify firewall options. And I can specify who or
which specific IPs can connect, and things like that. But I make it easy to connect
over the Internet by doing this. SQL port, by default 1443,
SQL authentication. By default,
If you enable this option, we made the Windows administrator
also a SQL administrator. >> Okay. >> And this is useful if I want to
connect from a different domain or from inside of a domain. Like if I have a work group and
I want to connect to SQL Server, then I can use these credentials for
that. As far as configuration, so we have tweaked this experience
a little bit based on customers. The really easy experience, very
very easy to understand and follow. This is specifying just
a number of files and that you need for your storage. So on-premises,
everybody understands today, IOPs, how many operations per
second I can do on the storage. And how many megabytes
are utilized per second. I need to write into
the storage right. We just exposed that and
if you have that information for SQL server which is very easy
to go to from for example. >> Yep
>> You know how many items you need. >> You can simply just go and find
the number of files you need here. And it takes these numbers based
on the size of the VM that you. >> Correct. >> Okay.
>> Correct, like in this case this
virtual machine, DS 14 allows me to go all the way to-
>> [CROSSTALK]. >> if I need less than that
I can go to other machine. If I need more than that I
can go to a bigger machine. >> Okay.
>> Right and obviously I can specify all
different number of disks that I want to be configured for
having more size. >> Okay. >> If I want all the way to 32
terabytes for example of space, I can specify that. So in this case, I'm gonna go small. Let's go to maybe 30,000 IOPS,
something like that. >> Okay. >> All right, and then it tells me
how many data disks are gonna be created as part of this experience. >> Okay. >> So in the end,
it's gonna create six data disks. Those are gonna be one TB each. >> Mm-hm.
>> All right, and all those disks are gonna
available to SQL Server to provide the right number
of IOPS and then to train it. >> Okay.
>> And finally, I can specify the optimization
of the storage. So this is a very
cool thing as well. What do you wanna do your work? So you want to do transaction
of processing for example, and you want your transactions to
commit as soon as possible? Then you probably want to use
transactional processing because in this case, what we're gonna do is,
we're gonna create a storage of space in
windows, it's just virtual disk. >> [LAUGH].
>> It's like a geological disk on top of on the disk. >> Hm-mm.
>> I want to create a write test stripe size and the
stripe size is, what is the size of the junk the sequel is going to
write into that virtual machine? >> Okay.
>> So you want to use transaction processing where you
use 64 gigabytes which is the recommended
size of those junks, right. >> So it does all of that for you. >> It does all of that for you. >> Absolutely, okay.
>> You don't have to think about, you how in the past, you went and read the best performance
practice for SQL Server. And you have to go, and
optimize your storage, and specify some transfer of SQL Server,
and all that. All that is done for you. >> Man, that's sweet. >> So you specify W or
the specify data warehousing. If all you care about is just being
able to execute complex ways, to read as much data as
they can from this data. >> So one is more read heavy
data ware is more read heavy, whereas transactional is
more read write heavy. >> Yes, it's basically,
do you want to complete a section as soon as is possible or do you
want to read as much as you can? And you don't care as much
about it [INAUDIBLE]. >> Okay. >> So let's say we want processing,
all right. The next thing is automated
batching, all right. So this is another common
thing that people have to do. So if you have a work load
that needs to run 24/7, would you still want to patch it? You'll want to patch it. You'll want to patch Windows, and
SQL, for the latest performance, for security hot fixes. You need to somehow specify when
you want the patching to happen. You can do it manually,
which is quite a bit of work, or you can do it through Azure. And here, you can specify when
do you want patching to happen, you can specify any day of the week, because this is by when do you
want the patching to start. >> Okay.
>> By the first two in the morning. This is UTC time, so whatever
time the machine is running in. This is gonna be two in the morning. >> Okay, so this is UTC time, okay. >> It's your time, your local time. >> Your local time, okay. >> Right.
>> Okay. >> So if you change [CROSSTALK]. >> Pacific coast time,
West coast time here. >> Right, so really like this will
be, this happens to be in Azure for example, at two in
the morning we shut down. >> Okay.
>> And then the maintenance window. And this how much time do you
want to dedicate to patching? So it can go all the way
to three hours there. What happens is we start patching,
we download all the patches, we start applying them. And we'll reboot the machine if
the patches require the machine to be rebooted. However, let's say that we already
heated the maintenance Window and we don't have anymore time, so time
we stop the patching, all right? We're not gonna do it anymore,
we pause it there. And then the next time,
the next week, when we heat again the patching
Window we're gonna continue. >> Okay.
>> And this guarantees that you have always
the latest [INAUDIBLE] but also that your work
is going to impact. >> Okay, good. >> So
definitely very useful to have. Then we have automated backup. Another very, very common
thing that DVAs have to do. We can enable it by default,
and what this will do, it will basically automatically
take backups for you, right. Now, let's go to the options. So the first option is
retention period, and this specifies how long do you want
Azure to keep your backups for? And also customers,
I only need them for a week. Some customers,
I need them for a month. So we can specify that in here. And what this will do is it
will keep those records for that period of time. What if at some point, we pass our
attention period in the back of another if we don't require them
anymore, we can delete them for you. And what this is going
to allow you is reduce the storage price cuz you're
not using the storage anymore. >> Yeah. What if you need them
longer than that, you just back them
up from the storage? >> Yeah, that's something
that we're looking into. Basically automatically support
long term retention and being able to specify up to years,
potentially seven years or so. >> Okay. >> Right now, the only way to do
it is using a script for example. The automatically looks at this at
the end of the retention period and copies this into a cold storage for
example. So now we have cold storage,
which is very, very, very cheap and you can store
years of data for example there. >> Okay.
>> So today still requires
the Power Shell script to do, is something that we're
going to have sometime soon. Encryption. So if you want to encrypt
your backups for security reasons and by doing this,
what will happen is, SQL Server will generate an encryption key,
and an encryption certificate. And that will be used
to encrypt my backups. >> Okay.
>> So obviously, your storage accounts
is already encrypted. And only you and your virtual machine have access
to that storage account but this is additional
security on top of that. >> Yep.
>> So your backup itself, is gonna be encrypted only
SQL Server will have. The ability to create a backup. A couple of new features, so now we
allow you backup system databases. So. >> There is, okay. >> We back up now Master,
Model, and MSDB. So it's still important
to have Master and MSDB. >> Hm-mm.
>> There is some instance level functionality that
is stored there for example, logins or things like Asian jobs for
example, or reputation jobs and
things like that, so by enable these basically we'll
back up [INAUDIBLE] and databases so that you'll have the guarantee that
you'll see some databases, and also backup, and you can always
restore it to some point in time. >> Yeah. >> There's another option, so by default we specify the backup
as scheduled automated. What this means is,
we'll automatically take backups for you and we'll take them based on the
amount of load that your databases generate. >> okay. >> So if your instance, your SQL
Server instance is generating a lot of load because it's very,
very active and you're doing a lot of index rebuilds, and doing a bunch
of bulk loads and things like that. And we're gonna get
it a little logs. So we're gonna take
backups more frequently. >> [INAUDIBLE], okay
>> And that works for some people. Some people don't want to have
to think about the schedules and how to do backups, and
it works great for them. Some of other customers
want more control. They want to be able to
specify how often do I want my full backups to do. >> Yeah.
>> For example. So now we allow them to do that. So you can say, I want to take my
back ups either daily or weekly. >> Hm-mm.
>> All right, these are full back ups. You can specify similar to
a patching Window when the back up will take place. >> Hm-mm. >> All right. >> You can also specify the time
Window for the backup, right? So during that time Window,
we'll go and try to backup all of the databases
in the server, right? If at some point, we can't finish because let's
say you have 100 databases and you only specified two hours, we'll
continue in the next Window as well. So I can specify
the Window here as well. And finally, the frequency
of your log backups, right? And this is how often in minutes
we're gonna take these logs backup, right? A couple things to mention, and others things just good opportunity
to explain to the audience. In Azure we have, we maintain
three copies of every disc for Virtual Machines, right? We have a continuous process
that is looking at the disks and we detect that for some reason
there is corruption in the disk. We verify a check somewhere,
we take that. [INAUDIBLE] Decommission
the disk and bring a new copy. So at all times we have three copies
and what that guarantees is that your data will always be safe, and
secure, and we have three copies, and we'll never ever,
ever lose your data. >> Okay. >> So you can see how this is
different from on-premises, where you have to worry about data. Because of that, you're gonna have
to take backups more frequently, because you don't know when
the storage system may go wrong. So in this case,
what is the purpose to the backup? Well the purpose of the backups is
really one for compliance reasons. Some customers required to have
backups up to a certain point in time, and being able to go back in
time in case it's required, and also in case of human error. So let's say that by mistake I
delete a table or I run some job that I didn't want to and suddenly
I updated a table by mistake, and I need to go back in time that's what
my backup are gonna be useful for. >> Hm-mm, okay,
[CROSSTALK] that makes sense. >> Data protection power you're
safe, you have three copies, your data will never been lost. >> Yep.
>> All right. So for the back up part. >> For the flexibility there, yeah. >> Yeah, so you can see a lot
of stuff we're doing here. All right, so cable we're gonna
touch on this one only for a few seconds. So again, for encryption, so
if you're encrypting your backups or it encrypts your data base through
[INAUDIBLE] database encryption. SQL will need to create a key, a
private key for database encryption. By default SQL stores the key
locally as part of the instance. We have something called the Azure
Key Vault and the Azure Key Vault is a secure HSM, it's a hardware store
module, maybe wrong with that one. So basically it's our solution
that allows you to store keys and manage keys in a secure way. And it has a secure channel between
the set of [INAUDIBLE] we have in Azure, and the module itself. And this allows any software dept
that needs to manage keys for encryption, decryption or key rotation to use that
module to store the keys. >> Gotcha, okay. >> It's just a secure
way of doing this. For people who are interested for
looking into cable. Today you are going to create cable
we don't have a US experience, yet even though we are working
into adding that to the portal. You can do it through power cell for
example. And once you have
the cable data URL for example, and some sequence you can
connect this to your SQL machine. And then SQL can use the cable
to sort of keys based on something else that. Okay. >> It's very interesting work. We're going to keep
improving the UI. That is something that is
available already today. Final one, our services. So starting with SQL 2016 we
have our services which is in the database of analytics and
this allows you to run one of hundreds of different algorithms for
things like forecasting, classification, clustering,
basically the principles of Machine Learning,
integrated within SQL server. So there's a lot of interest
right now in our language and our services and. >> Yeah,
we just had UC on last week. [INAUDIBLE]
>> Good, good. >> So, good.
>> So if you want to go and try our services there is no easiest
way to do that than going onto virtual machine with SQL developer. Right and then you can
enable our services here. Which basically is going to download
the instead of you know some of this things are open source so
we gonna download them. We are going to configure them,
install them. By the time that this machine is
ready and running you can just go on automatically and start running
your on top of SQL server >> Okay cool. >> So we are going to this
experience when we click okay. We're gonna see this summary of
all the things that we have done. So you see we have the name
of the virtual machine. We have this SSD. We know that it's a 16 core machine. Public connectivity for SQL Server. We automated, so
we enabled SQL authentication so we can authenticate to
SQL Server using my credentials. We enabled our services. We optimized this machine for
transactional processing, right? For six terabytes of storage, for
8,000 768 megabytes per second. We ultimated patching on
Sundays at 2:00 in morning. We ultimated backups. We keep our back ups for 30 days. We encrypt the backups the back
ups we are going to be taking every week at 2:00 in the morning
we are going to give six hours to the backups we're gonna
log back up every 60 minutes. So see all the stuff that we do. >> [LAUGH]
>> I mean obviously we went through this and we explained every options
and that's why it took us so long. Once you understand you can probably
go and do this through two or three minutes. >> Yep.
>> Right? And your machine will be
running in ten minutes. So 12 to 15 minutes. In ten minutes you have a machine
with all of these thing. Can you imagine today on premises
how long will it take you to get a machine and computer and optimize
it and specify all of these options? >> It's assuming you
have the hardware. >> Assuming you have the hardware. Right. So this point you know, I could
just say okay and this will go and create my machine, again ten
minutes and it will be running. I could also download a template and the template is going to show me,
a JSON template. This is a declarative way that it
specifies all the arguments that I have specified through
that configuration. >> Wow.
>> And usually there is a lot of stuff here. You know you can open
this to Virtual Studio. And then go and modify it. I mean you could open it
in Notepad if you want. >> Yeah [LAUGH]
>> Probably be a little easier. And you see all the arguments here, and the purpose here is
you can download this, and after you download you can
add it to your library. And then you can go and replicate
this 100 or 1,000 or 10,000 times. >> [INAUDIBLE] automate
the deployment of my, wow okay. >> You could just run
this in a loop, an so you have 100 of these virtual
machines with all the we specify. >> Okay, very cool. That's amazing. >> So you can do that,
you can click OK. And now it's gonna go and
create the machine. It's gonna take
a little bit of time. So, we're not going to wait right
now, cuz we need to finish the show. >> [LAUGH]
>> So we're gonna show one
that is already running. So here we have one that
is on in SQL 16, and the presentation, is our diagnostics
I was mentioning before. So we show things like CPU, and
information about disk reads and writes, information
about the network, etc. I can go and change anything that I did before,
that's an important point. Don't worry about-
>> [LAUGH] >> If you select something running in the beginning, I can go and
select SQL Server configuration, for example. And I can go and
change anything that I want, right? Like I can go into storage for
example, and I can add new drives or I can extend my drive, very,
very easy to do, right? I can go and change
the connectivity, the patching, the backups, our services. [CROSSTALK] So
everything can be changed. >> So I don't have to RDP
into the box, and change it? I can just do it right here,
and it'll make the changes. >> You can do it. Right. It's so easy. >> Yeah. >> For people doing development and
testing, sometimes I hear some customers say,
I started my trail and I run out of my credits
sooner than I expected. You can go on configure
auto shutdown, and auto shutdown you can specify
if you want your machine to be automatically shut down
at the end of the day. So if you come to the office at
depending on what time you started [INAUDIBLE] at ten. >> [LAUGH]
>> Probably work on the evenings, but you just specify
whatever time you want. And you can specify when you want to
shutdown your machine automatically. So you can specify
7:00 in the evening or whatever time you want and your
machine will stop at that time and you will stop paying for
the compute cost of the machine. >> That's awesome and
will it automatically start at that? Do you I have to
manually restart it or will it just shut down and
restart for me? >> So today, you still have to
manually restart it in the morning. Because we don't know if the gonna
be at 7:00 in the morning, or at 1:00 in the afternoon
>> [LAUGH] >> Depending on whatever they were doing on the previous night. So today's planning actually
to show that as well here. But right now we
have auto shut down, so again a great way to say compute-
>> Yeah save me some money because you know
if I'm, I go home at 7:00, come back in at 7:00, 8:00 or
9:00 and they'll come in until10:00. I don't want to be paying for those hours, the compute hours
when I'm not using the box. >> Totally and this could, just imagine instead of running
the machine 24 hours in a day, you're running eight
hours as one fourth. So you can extend three
times your utilization. >> Yeah, we do that all the time. I ran out of credit. >> Yes that's a challenge. If you're doing a good job and
your computer auto shut down for example, you can go and use SQL
developer which again is a free license SQL server on a four core
machine for a whole month for development and
testing we had a trial account. >> Wow.
>> A whole month for free. >> Yeah.
So sign up now, why haven't you signed up? >> Go on and do it, right? >> [LAUGH] Yeah. >> All right, so
I think that's all I want to show. Obviously, you know, what happens
once you go and do all that you will have an IP associated
with the machine or a DNS name. Like in this case I haven't
created a DNS name yet. When I have an IP here so I can just
go into my SQL server management studio on my laptop for example and
I can go and just connect. Connect my engine connect to
that IP address, or DNS name and I have a SQL server that is
running and I can just go and do whatever I want to do. >> And I don't have to do
anything other than that right? Just tell the public and it will go
create the firewall rules for me and the OS things like that. >> Yeah correct. You can specify whatever you want,
right? So anyway. I hope I have my skills at work. >> [LAUGH]
>> Okay. So you can go into that right and do whatever you want.
>> Very nice. >> So yeah that is basically the experience. We continuously keep improving. This is something that is
only available in Azure, you won't find anywhere else. Anywhere you'll find just
a virtual machine and you'll have to do everything else. >> Everything else. Great experience and I love that
cuz there is, it proves that we're continually thinking about how do we
make the experience better not only for, including how do we save you
money which is awesome, right? >> Yep. >> Because You know everyone's
like we're not gonna charge you. Well no we want to save you money. >> Yep.
>> Right. Plus make the experience better. Very nice. I'd love to have you back for
more of this and I love the always on one
because that'd be a good one. >> This was awesome. [CROSSTALK]
>> Hey. Awesome, thank you,
thanks for coming. Hey everybody, thanks for watching, we look forward to having Luz back
and we will see you next time. [MUSIC]
