How failover works in SQL Server Clustering between two nodes

okay so here we are going to see how does the cluster serves the purpose of high irritability like if there are two Knorr participated in the cluster how does it ensure that one of the Knorr when goes down or for envision was offline how does the other node takes out the services and keeps the keeps the sequel's a state a leave so yeah we can just go and see that here I have for virtual machines connected in my laptop here created my laptop this is the domain controller nothing else I do this is the domain controller nothing to it this this is the static domain I can you take these three mowers to this domain but I will configure failover cluster between these two nodes dev anyone in the event well just go and show you all this right here so you can see here this is the failover cluster manager you can see the status of the cluster here the name of the class which they plaster and it is showing like we have two clustered roles and two nodes participating in this cluster you can see the roles the mr. TC service and the sequel server service both are currently by Dave and one this is the virtual server named sequel server name and these are the supportive this is the data disc and this is the log disk and these are the dependency sequencer innocent so are currently hosted by n1 and you can see both the nodes are up here right now and in the same is the case for all that exhausted by the n1 this the Dirac Slovaks this quorum the heart of cluster and these are not configured it for any use so these are the networks the private and the public need for the power serving the purpose of heartbeat connection between the nodes and other one for the communication to the outside network for the third parties for their pics and string connecting from other sources so you can see our clashes very early there are no readers no failures no warnings so but that's a very very very good configuration you can you can see the question of the cluster you can manage the cluster from the other node as well you can see from here as well the same thing but what have these are participating in the cluster so you'll just probably see the same thing from the other node they are n - you can see both the service is running by n1 at this time both the nodes are the storage disks are hosted by n1 at this time so we can we can we can go and try to fill our or or first of all let me show you how to connect to the sequel server we actually the name of the server is Dave n1 which smart also but the virtual name of the sequel services here either so so so we have to make sure the virtual name in or the name of the server so let's try to connect so we should keep the virtual name here and see okay see so you see that it's connected right so be up this what I wanted to show you now let's go and try to do the failover let's go and try to put the manual fail eventually to and to both the services if spirity fast because it has not a lot of dependencies depending on the IP address so it would not take that long but this is dependent on the virtual name the virtual IP address the server's the data file of phyleus right and yes the sequel server and is it as well you can see they have all failed to end to the dick should also be felt when to you can see here both the noodle is still up these two dicks are already failed of trying to and it's not necessary to file these as anyway we can feel that as well select no because it's not the dependency right now so anyway so currently they are now assured by now and to know what happens if for any reason and to hose down you can see it as well all of the roles are currently now fostered by n - Donna Logan - 104 let's say for indigent end to go stop we can go in restart we start this this server so we can see this in the demo here okay let's go power restart mist restart now come to the n1 and see currently is hosted by n2 right so how does it get affected okay see so you can see now right anyone just starting to host servers now and making sure you don't have any any time doubt or like the application doesn't get no connection masses so you can see now the services are back online so how fast it is it actually depends on the quorum setting saying the network resources the network as well so you can see we are still connected to the sequel server so I have any any issues connecting the sequel server I don't have anything right now so this is what I wanted to show you this is called actually high ability even if one server goes down you can see they went to is down but still the sequel services are up and running we don't have any problems for the applications all the users are back online in fact they are they're always online maybe they will get the taxol milliseconds or even the seconds of downtime but they'll not even experience that right so this is how cluster serves the purpose of hive ability even if one of the server goes down the other server takes out the services back and the services keeps them so yep okay so the service is coming back online so either we don't have any problem the service the services is up and the sequel services are running we can see the state of okay so we have fun activity now if the saying was removed from the active in our cluster though and to the cluster service on this node may have a stop or whatever okay so yeah this is this is what we wanted to have one event right because anyway this end to us down you can see in the status of node here okay I think this is already up now oh yeah this is already up so now you the cluster is able to ping that server so anyway this is what I wanted to show you hope this would be very helpful thanks guys please let me know if you fail this really helped land for any solutions thanks guys

Channel: Ashok Deo

Views: 19,744

Keywords: Computer Cluster (Field Of Study), Microsoft SQL Server (Software)

Id: 8LRrFPSZGbw

Length: 7min 50sec (470 seconds)

Published: Wed Oct 14 2015