0:01 foreign [Music]
0:05 [Music]
0:07 in this video we are going to learn
0:09 about what are managed instance groups
0:11 how are they different from unmanaged
0:13 instance groups we will also look at the
0:15 features of make offers which is auto
0:18 scaling Auto healing load balancer we
0:21 will do a live Hanson lab where we will
0:23 see all these things in action not just
0:26 theoretical part this video is very very
0:27 important make sure you watch it
0:30 completely because whatever Services you
0:32 use on Google Cloud whether it's
0:35 kubernetes or bigquery or any other
0:38 product most of the services backbone is
0:40 Mig so if you understand Mig is really
0:43 well your life is really easy while you
0:45 work on your projects so basically
0:47 compute engine offers two kind of
0:49 managed and unmanaged so far in the
0:51 course we have been using unmanaged
0:52 instance groups first of all let's try
0:54 to understand what is the difference
0:56 between managed and unmanaged Minister
0:59 scrubs unmanaged instance groups lets
1:01 you load balance across a fleet of VM
1:04 that you manage yourself which means we
1:07 have to manage our instances what is the
1:09 size or how many times you want to keep
1:11 it up if you have to add servers if you
1:14 delete servers everything is under our control
1:15 control
1:18 right Google did not help us to manage
1:20 those things however we'll talk about
1:23 managed instance groups managed instance
1:25 groups are basically a collection of
1:27 virtual machines that can be used as a
1:28 single entity
1:31 and its instance groups let you operate
1:34 applications on multiple identical VMS
1:35 we can ensure that our applications
1:38 hosted on Mig instances are highly
1:40 available we can also take advantage of
1:42 automated mixed services like Auto
1:44 scaling Auto healing Regional
1:47 multi-reason multi-zone deployments and
1:48 automatic updating so these Concepts
1:50 like Auto healing Auto scaling auto
1:53 updating these things are you know not
1:54 something which you use day in day out
1:57 so let's try to understand them one by
2:00 one what is auto healing whenever a
2:02 service goes down the the VM will
2:06 restart itself okay the VM will kill
2:08 itself without any manual intervention
2:11 it will restart itself and start serving
2:13 the traffic how to configure that we'll
2:17 see this in this video itself second is
2:20 multi-zone so you can have your one
2:22 instance deployed in your Center one one
2:24 instance in U.S east one so that's how
2:26 you can distribute your workload across
2:29 different zones within one application
2:31 what is load balancing you can have
2:33 multiple instances in different zones
2:35 and you can distribute the traffic while
2:37 load balancers the fourth bit is around
2:40 Auto scaling so whenever something
2:42 whenever a load Peak load happens you
2:44 need to ensure that you're increasing
2:46 the size of your Fleet which is your
2:48 number of VMS should increase to
2:50 accommodate the traffic
2:52 that's where Auto scaling comes in we'll
2:55 try to see that auto scaling also in
2:57 action the last bit is auto updating so
2:58 whenever we have some safe type of
3:01 updates auto updating helps us to update
3:04 update our operating system patch any of
3:05 those things are done
3:08 make instances are ideal for stateless
3:11 survey for all sort of batch workloads
3:13 or stateful applications
3:15 that's the basic difference between
3:17 managed and unmanaged instance group you
3:20 can also deploy new versions without any
3:21 downtime under under management
3:23 subscript which is making instances
3:26 before we move to understand the
3:28 management stress group in detail let's
3:30 try to understand what is auto scaling
3:32 so suppose one of the examples in terms
3:35 of this IPL season today IPL is
3:38 happening okay so whenever in your first
3:41 and second over suppose you are able to
3:45 serve with two servers or entire traffic
3:47 the in the third over what happens Ms
3:49 dhoni or maybe a Virat Kohli comes to
3:52 bat suddenly the number of users will
3:54 increase and you'll have to add two or
3:57 three more servers to address the peak
4:00 load when they get out or you know the
4:02 match is finished the traffic will
4:04 reduce gracefully
4:05 and then you have to shut down your
4:08 servers to come back to the same load
4:09 which you are able to serve when you
4:11 started the map that's the basic graph
4:14 of Auto scaling as your demand grows you
4:16 have to scale your VMS and as your
4:19 demand falls down you scale down your
4:21 VMS right that's what Auto scaling
4:23 actually looks like that's one of the
4:24 example I thought to give you now let's
4:26 go ahead and understand what it takes to
4:29 create a management sense group the most
4:32 basic part is instance template if you
4:34 have an instance template you can create
4:36 a medicine instance group you can also
4:37 create it unmanageable what is instance
4:40 template instance template is basically
4:42 a resource that we can use to create
4:44 virtual machines instance templates are
4:46 very very convenient way to save a VM
4:48 configuration and then we can later
4:50 create VMS group of VMS or reservations
4:53 later on why because it gives us a
4:55 template and when we have to Auto scale
4:57 we have everything ready we just have to
5:00 create one replica of it and it's not us
5:01 we will configure it Google will create
5:03 it for you that's the benefit of using
5:06 make instances so instance template
5:07 basically will Define what is the series
5:09 and machine type you're using service
5:12 account your boot disk OS image if you
5:14 want to run a container image just pass
5:16 that the labels the startup script Network
5:18 Network
5:20 everything right so you just figure out
5:23 uh you just said that you set all these
5:26 details in in your templates and it will
5:28 boot up your disk it will boot up your
5:31 VM wherever you uh Auto scale or descale
5:33 or when you create it so that's
5:36 templates serves as a read-only copy
5:38 very important concept instance
5:39 templates are designed to create
5:41 instances with identical configuration
5:43 so you cannot update the existing
5:45 instance templates or change any
5:47 template after you create it so it
5:50 becomes a read-only copy so you just
5:52 have to create another version if you
5:54 want to make any changes okay that's
5:56 very important and we you cannot go
5:58 ahead and change an instance template
6:00 once it is submitted to understand this
6:02 we are going to do a lab which has these
6:05 steps the first step is to create an
6:07 instance template and to deploy nginx
6:09 with all required properties in multi reasons
6:10 reasons
6:12 then we'll be creating an auto scale
6:15 link then we will be creating an auto
6:16 scale manage instance group using that
6:18 same instance template which we create
6:21 in the first step third step is around
6:24 creating a health check than creating a
6:27 load balancer stopping the nginx to see
6:29 auto healing in action increasing the
6:31 road to see auto scale interactions
6:33 right so this is the architecture we are
6:34 going to do in lab
6:36 see you at the labs
6:38 first of all we have to create an
6:40 instance template to deploy nginx with
6:42 all required properties in multi-ts so
6:44 what we'll do we'll go ahead and create
6:45 an instance
6:48 for that we'll come to this we'll come
6:50 to the console and click on compute engine
6:52 engine
6:54 over here you'll see instance template
6:56 as I mentioned we have to create an
6:58 instance template click on instance
6:59 template click on create instance template
7:01 template
7:04 you can give a meaningful name say cloud Sprint
7:06 Sprint template
7:13 in series N1 and give me F1 micro which
7:16 which is the most cheapest one
7:18 okay I'll go ahead if you want to deploy
7:21 a container with every VM you can pass a
7:23 container over here but in this case we
7:25 just want to use the startup script so
7:27 I'm not putting any container name here
7:30 I'm just using Debian as a image
7:34 I'm not changing any of I'm not changing
7:35 the service account just keeping a
7:36 default one but if you want to choose
7:37 you can choose here
7:40 everything is by default I am going to
7:42 allow HTTP in HTTPS next is Advanced
7:44 option now network is I'm putting it as
7:47 default what I want to add is uh
7:49 Management in here I want to pass a
7:52 startup script which will install nginx
7:54 and set it up for me for that I'm just
7:55 going to use this script but it will do
7:57 it will first of all update the voice it
7:59 will install nginx start the service and
8:01 then it will it will replace this HTML
8:03 file with the host name so we will know
8:05 that which host name we are being served
8:06 the traffic from
8:08 nothing else I don't want to change
8:11 anything it should be standard
8:13 let's click on create so this will
8:15 create an instance template very very
8:18 quickly for me okay once the template is
8:19 created we can move to the Second Step
8:21 meanwhile this template is getting
8:23 created we will go ahead and check what
8:26 is the second second step is we need to
8:28 create an auto skilled managed instance
8:30 group using the same instance template
8:32 now you can see this instance template
8:34 is created with Cloud Sprint template
8:37 name machine Type image name disk and
8:39 everything else is here
8:41 you can create a virtual machine from
8:43 here or you can create an instance group
8:46 from here okay let's go to instance
8:47 groups we don't have any instance group
8:50 existing uh for for now the Second Step
8:52 was to create an instance group using
8:54 the same template so click on create
8:56 instance scope you can see here many
8:59 type of options you have first is
9:01 unmanaged instance Group which is uh
9:03 we're not going to use in this case we
9:05 are going to use the managed one then
9:06 there are two type within managed it's
9:09 stateless and stateful stateless is uh
9:11 basically for serving and batch
9:13 processing stateful is used for
9:15 databases or Legacy applications I have
9:18 to choose the instance template now okay
9:19 which we just created Cloud Sprint
9:22 template what I'm going to do is just
9:25 let it be no changes now the requirement
9:27 was I have to create this in multi-zones
9:29 so what is the meaning of multi-zone
9:31 suppose you have customers from
9:34 different different areas within us you
9:36 can choose to serve a few of your
9:39 servers can be in one C area few of your
9:42 servers can be one F zone so you can
9:44 serve the customers from all these areas
9:46 in a better way to avoid latency they
9:49 can have the better user experience so
9:50 I'm going to choose these two you can
9:53 choose more if you need and target
9:56 distribution shape is even or balanced
9:59 so I'll just put even for now
10:03 okay next step is auto scaling in a
10:04 nutshell Auto scaling is all about
10:06 automatically adding and removing
10:08 instances to the group for a period of
10:11 high and lows but what I'm going to say
10:14 I need two minimum instances and maximum four
10:15 four
10:18 okay a minimum means that every time
10:20 these number of instances will be
10:23 running maximum is irrespective of
10:26 whatever the load is I'm I'm not asking
10:29 you to create more than four VMS while
10:30 working with free tier you might get
10:33 issues when you try to create more than
10:34 your quota which is by default I think
10:38 it so try to test it within that range itself
10:39 itself
10:42 now when you want the auto scaling to
10:45 happen you want signal right it will
10:47 work on an event so by default it is if
10:50 CPU relation is more than 60 percent it
10:53 will create another instance now what is
10:57 Target CPU utilization 60 50 40 80 you
10:59 can choose depending on your application
11:02 how much time your new node is taking uh
11:04 to be ready you can choose depending on
11:06 that okay click on done if if I want to
11:09 add another signal I can choose HTTP
11:12 load balancing you need if any based on
11:14 some event I want to add suppose I want
11:17 that when an e somebody uploads a file a
11:20 VM should get added to to this node pool
11:23 to create to process that request so
11:25 that I can do here any monitoring
11:26 metrics I can check here so you can add
11:28 different type of singles in this
11:31 example we are using CPU utilization
11:33 then you have cooldown period this is
11:35 very very important concept uh this
11:37 basically specifies that how long it is
11:41 going to take your app to initialize
11:43 okay how long it is going to take your
11:45 app to initialize from boot up time
11:47 until it is ready to serve what happens
11:50 you add a server okay you add a server
11:52 and you are still not able to serve the
11:56 traffic from that that VM because your
11:58 application is not set up properly
12:00 that will give a 404 error and that's a
12:02 bad user experience so you should ensure
12:05 once your website uh once your server is
12:07 up and ready it's already boot up how
12:09 much time it will re it will take your
12:11 app to initialize in this case it is
12:14 nginx all right so um in this case we
12:16 are saying 60 second I believe 60 second
12:18 is more than enough to uh for nginx to
12:21 set it up and get started next is about
12:24 keeping the instance life cycle I'm not
12:27 going to change anything yep next bit is
12:28 around Auto healing
12:30 what is auto healing Auto healing will
12:32 recreate VM instances if an application
12:34 cannot be reached by the health check so
12:35 health check is an entry let's create a
12:37 new health check
12:39 say cloud
12:43 Sprint let's see what it will do it will
12:45 it's send
12:48 request every seconds whatever you
12:49 configure so in this case the criteria
12:51 we're going to find is check intern
12:53 values every five second timeout is five
12:55 second healthy threshold two consecutive
12:57 success is unhealthy threshold two
12:59 consecutive success so basically if
13:02 twice it will try two times okay if your
13:04 your service is not responding it will
13:07 Mark your instance as unhealthy and
13:10 based on that kpi you your instance will
13:12 be recreated that is the benefit of Auto
13:14 healing so if if this kind of situation
13:17 is happening at 1am or 3 am nobody have
13:19 to sit manually and you know recover
13:21 that this Auto healing feature will take
13:23 care of your instance it will remove the
13:25 node which is not able to serve and
13:27 create a new one that's the benefit of
13:30 Auto healing again we have 300 seconds
13:33 of initial delay to ensure that it is
13:36 warmed up we just chose the template the location
13:38 location
13:40 Auto scaling policies
13:44 and sixty percent Auto healing policies
13:46 Auto healing policies and Port 80 and
13:50 I'm going to create the instance group
13:52 it will take some time to get created
13:54 I'll be back once it is ready
13:57 okay so within two minutes the instance
13:59 is created you can see we had requested
14:02 minimum two number of instances it was
14:04 created at this time the instance is
14:07 ready template we used is this and
14:10 instance are healthy we have used this
14:13 CS health check for uh for the health check
14:14 check
14:16 for auto healing basically and for auto
14:18 scaling we have said that whenever CPU
14:20 utilization is more than 60 just scale
14:23 one uh add one more node and maximum can
14:25 be four so these are configurations we
14:27 did while uh you know installing it
14:29 let's go ahead and check if our nginx is
14:36 okay you can see uh you are being served
14:38 from Google welcome to Google Cloud
14:40 platform you're served from this
14:42 instance Group which is uh
14:45 7p67 right 7p67
14:51 this
14:54 you'll be you'll get that you're being
14:59 served from 641 641 okay why uh I change
15:01 this script like this because when you
15:03 use use load balancer you will be able
15:05 to see this in action that how the
15:08 traffic is balanced between uh two
15:10 different VMS deployed in different
15:12 zones like this is deployed in C this is
15:14 deployed in F so when you get customers
15:16 from C this will serve the request when
15:18 you get from this Zone this will solve
15:20 request so they can have a better user
15:21 experience that's the benefit of having
15:24 multi-zone deployments to have a highly
15:26 available application coming back to
15:27 here we have created an instance
15:30 template in multi-reason installed nginx
15:32 Auto scaled manage instance group using
15:34 the instance template the third step was
15:36 to create the health check for auto
15:38 healing we also did that the fourth step
15:40 is to create a load balancer to
15:42 distribute the traffic among VMS added
15:45 this is the architecture diagram that's
15:47 what we have to do so we have we are
15:49 done with the right bit where we have
15:51 set up the managed instance group now we
15:52 have to create a load balancer so
15:54 customers can directly reach to this
15:56 reload balancer and we can serve them
15:59 from these different VMS hosted in
16:00 different zones go ahead and create a
16:02 load balancer to create a load balancer
16:04 we need to go to network Services load balancing
16:06 balancing
16:07 load balancing create a load balancer
16:10 here you will get three type of load
16:11 balancer while we study load balancers
16:14 in detail I'll be you know showcasing
16:16 all the examples but just for now you
16:18 can understand we'll be using https
16:20 because this is mostly used for internet
16:21 facing or internal single or
16:24 multi-regional kind of uh application
16:26 click on start configuration to create one
16:27 one
16:30 or the first thing is internet to my VM
16:32 yes I the second option could be between
16:36 my VMS but I need from internet to my VM
16:38 next is global HTTP load balancer
16:40 classic uh the classic one is not having
16:42 the traffic redirected rules so we
16:44 should use the newer one which we may be
16:46 we never know we might need the traffic
16:48 redirect the rule as well in future
16:50 click on continue your select a name
16:52 front end protocol
16:55 Port will be 80 done
16:58 okay we have added front-end ipn port
17:00 next is back-end configurations in
17:02 backend configurations we have to create
17:04 a backend service
17:05 let's click on create a backend service
17:11 I'll say cloud CS backend service
17:16 backend type can be instance group it
17:18 can be serverless it can be private
17:21 service connect and hybrid these all are
17:24 big options okay you can you can choose
17:26 as you need tomorrow when you
17:28 extensively work you can choose them
17:30 back and type in this example I'm going
17:32 to choose instance groups HTTP timeout
17:34 is 30 seconds
17:36 new instance group so this is the
17:37 instance Group which we have created I'm
17:41 saying port number is 80.
17:44 port number is 80 okay I don't want CDN
17:46 in this case
17:48 health check I'm saying okay yes I have
17:51 health check I'll use CS health check
17:53 no security policy for now click on create
17:54 create
17:55 say okay
17:57 so I have said that my front end
18:00 configuration is there as a backend
18:02 configuration I have added my instance
18:04 group so whatever I have hosted on my
18:06 instance group that can be served here
18:09 I'll give a name and I click on create
18:12 so this will create a load balancer for
18:15 me which has these two VMS added as
18:17 their backend alright so this load
18:19 balancer creation is completed we'll
18:22 open this and see if we are able to open
18:24 the website
18:27 okay perfect so we are being served from
18:31 this instance 7p 67.
18:35 okay so you can see that uh we are being
18:39 served from 7p67 this one okay this
18:43 completes the step four as well okay
18:45 you can see we have created a load
18:48 balancer hosted a compute Engine with
18:49 nginx we are able to create a health
18:52 check we'll see that also in action
18:55 now I'm going to I'm going to show you
18:58 two steps first is stopping the nginx
19:01 this will help us see auto healing in
19:05 action also how a load balancer helps us
19:07 to you know route the traffic to a
19:09 different node whenever something is not
19:11 healthy you can see now we are being
19:15 served from 7p 67 VM so I'll just log in
19:17 here SSH
19:22 okay I am here now what I need to do is
19:24 first of all what I'll do I'll stop the
19:26 nginx service
19:28 what I'm doing I'm stopping the nginx service
19:30 service [Music]
19:36 okay you can see it's stopped
19:39 you can see it's dead so we stopped the
19:42 nginx service which means
19:45 this particular VM which is uh which is
19:48 sp 67 is not able to serve the traffic
19:50 now two things will happen Auto healing
19:52 will come into the action first of all
19:54 see how load balancer will help us so so
19:56 far we are being served from this
19:58 particular VM you can see it's dead now
20:00 it's being recreated that is the benefit
20:02 of Auto healing so somehow your service
20:05 stopped working and you are not able to
20:07 solve the traffic from that VM so Auto
20:10 healing will check that okay instance is
20:12 not healthy this health check will check
20:15 it and recreate your VM that's the first
20:16 part of
20:19 Auto healing the second bit is now we
20:22 have only one VM running which means
20:24 that traffic which was being served from
20:26 this VM should shift to this VM that's
20:28 why we pay for load balancer let's go
20:31 ahead and check if that is happening enter
20:33 enter
20:36 yes you can see the traffic has shifted
20:38 to 641 VM which is the active and
20:40 healthy one
20:43 right that is the benefit of using load
20:45 balancers so load balancers
20:47 automatically the through health check
20:51 got to know that this 7p 67 is unhealthy
20:53 and I need to shift the traffic to 641
20:56 for a seamless migration that is the
20:58 benefit of using um using Auto healing
21:01 with load balancers
21:03 right I hope this you are able to follow this
21:04 this
21:07 now with auto healing this VM is being
21:09 recreated it will take approx two to
21:12 three minutes to get recreated set it up
21:13 depending how much delay we have given
21:15 and yeah you can also see this instance
21:18 health is down to 50 percent this VM is
21:21 recreated and this will be it will be up
21:23 and running once it is up and running it
21:24 will again start serving the traffic but
21:27 for now we will have to serve all the
21:30 traffic from this uh one six four one BM
21:33 because and this will be sort for entire
21:37 Zone unless this Zone uh VM unless this Zone
21:44 that's how you design a highly available
21:47 architecture even if your Zone goes down
21:50 you can still serve your customers you
21:53 can see now it is updating which means
21:55 which means this instance is being replaced
21:56 replaced
21:59 so far we have completed all these steps
22:01 the only step remaining is to increase
22:03 the load and see auto scaling in action
22:05 once we are recovered from the auto
22:07 healing once auto auto healing VM is
22:09 available we will see auto scaling in
22:12 action so you can see finally with this
22:14 it is showing that one VM is active
22:15 serving the traffic another one is not
22:18 healthy and it is being recreated
22:20 it is being recreated and it is being
22:23 updated with the new it is using the
22:24 same instance template all right so
22:28 perfect so it's ready to again serve our
22:30 traffic and it has recovered itself it
22:33 has healed itself let's go back and try
22:36 to access it six four one yes it is back
22:38 to sub backend serving the traffic
22:42 healthy okay that's the that's how Auto
22:44 healing works and that is how it is
22:46 going to save you a lot of midnights
22:48 where you can peacefully sleep and let
22:51 Auto healing take its new own course
22:54 and also you understood that how load
22:56 balancer will help you to you know
22:58 balance the traffic between these two or
23:02 more begins and uh if one in zone goes
23:03 down you can serve the traffic from
23:05 other Zone just by simple configurations
23:07 which we did in this video
23:09 the last bit in this video I want to
23:11 cover is auto scaling
23:13 to manage Auto scaling first of all
23:15 simply understand we have set maximum
23:17 four uh instances we can get added and
23:19 once our
23:21 CPU utilization goes beyond 60 then only
23:24 adds add VMS so what I'm going to do
23:27 since I don't have a live traffic I'm
23:29 going to simulate a traffic to simulate
23:31 the traffic what I'm going to do I'm
23:33 going to install stress it will
23:36 basically simulate a load for me let me
23:39 install it
23:41 okay it doesn't start I'm going to
23:44 increase the load this will
23:47 create simulator load and stress that
23:49 the CPU and it will increase the
23:52 significant timeout okay let me click
23:54 enter it let's see how it is increasing
23:55 the load
23:57 and there are many ways to see it you
24:01 can just go to this particular VM
24:03 what it will do it will increase your
24:05 load and how to see that you can click
24:07 on observability
24:09 and just see your instance increasing
24:13 then Note One Thing even if one VM
24:15 utilization is 100 overall utilization
24:18 has to increase then only Auto scaling
24:20 will work it's not like you have just
24:23 one VM running at 80 percent overall
24:25 it's still less than 60 it is not going
24:29 to add any servers so for overall
24:32 monitoring we'll come here okay perfect
24:35 perfect you can see now
24:37 you can see now stress command is
24:40 completed and it increased my load more
24:43 than 60 60 percent and I have two more
24:46 VMS added it is of course not in healthy
24:48 condition but it is being verified so I
24:50 have four instances Auto scaling is an
24:53 action kicked in initial delay which we
24:54 have put that is it is warming of the
24:57 instances and it will start serving the
24:58 traffic once it is available and ready
25:03 to be served right a simple load I just
25:04 increase the load and you can see auto
25:06 scaling happening in action without any
25:10 changes so okay you can see we have two
25:13 got two more friends to help us to you
25:15 know serve with this load and we are
25:18 able to quickly add two more servers go
25:21 ahead and see if we can get a service
25:24 from new VMS you can also check
25:26 the current load you can see the current
25:29 load has just is decreasing so the
25:31 moment overall utilization comes below
25:34 60 percent these two new nodes will go
25:36 we'll again come back to our minimum in
25:38 configuration which is two
25:40 that is what I explained at Auto scaling
25:42 uh happens so when you have load you add
25:44 the servers when the load is gone you
25:46 remove them automatically you don't do
25:48 anything just configure the kpis and
25:50 Google Cloud make instances will help
25:52 you to do that you can also see it is we
25:55 are being served from new VMS now okay
25:57 well one end was not there again one six
25:59 four one we are being balanced seven P
26:02 sixty seven so it is just shifting the
26:04 traffic to new new VMS and whatever is
26:06 where available it is being served from
26:07 there okay now I'm not going to do
26:09 anything I'll just wait for the load to
26:11 settle down once load is settled down
26:13 these two VMS should go away which we
26:16 just added as our Auto scaling I'll be
26:18 back once it scales down
26:20 okay now you can see the load has gone
26:23 down and it has started updating the VMS
26:24 so first thing it is doing it is
26:27 deleting the fourth wheel which was just
26:29 added as uh okay it started deleting
26:32 this one as well right so that's how
26:34 the moment the load got reduced it is
26:36 deleting these two VMS and we'll be left
26:38 with the minimum configuration which is
26:40 two right see those two instances have
26:42 been removed I mean we are going back to two
26:43 two
26:45 just to give you a recap we created
26:47 instance template
26:50 we created an instance group
26:53 we created a load balancer we created a
26:55 health check tested Auto overhealing we
26:57 created Auto scaling policies and we
26:59 have latest tested Auto scaling
27:03 hope this was an informative video and
27:04 you are going to go ahead and try your
27:07 hands on this lamp thanks for watching
27:10 all right so this concludes the big
27:12 instance group I hope you like the video
27:14 and if there's any question just please
27:16 drop a comment and let's try to discuss
27:18 them yeah thanks for watching see you in
27:20 the next one foreign
27:23 foreign [Music]
