2.2 Implementing DHCP Failover in Windows Server 2016 (Step by Step guide)

in the second part of the videos about implementing DHCP in Windows Server 2016 we are going to take a look into the high availability options for DHCP hello friends this is Nick from Innova solutions and today I'm going to continue with the second part of the three part videos about DGP and implementing DGP in Windows Server 2016 so in the first part we discussed how you can install on Dero how you can configure pretty much your DHCP server so it can start giving IP addresses to your clients but the next one and this is a quite important thing is to consider how your DHCP server can be highly available because this role is required and it's pretty much critical for all the clients within your network without any static assigned IP addresses and if this role if this row is not available on your network and it will make a disruption and the clients will not able to work properly not able to communicate between each other so we definitely need to think of a way to preserve to save this one so everyone can work without any disruptions so for the purpose of this video I would need some two servers and one client desktop machine the first server is my domain controller this is 0-1 the second server I'm going to use is my storage server but in this server I'm going to install the HP row so we can configure the settings for DGP fell over and I can show you how things are working but in your case you can decide to install the role on another domain controller if you wish or another separate server depending on what is your environment and how you want to protect your DHCP server and in the end I'm going to use my PC 0 1 so we can test the DTP fell over confirm that it's fully functional and able to still provide IP addresses if one of the servers is down so to start with DHCP high availability we need to take a look into some older methods that are still present in Windows Server 2016 and you can configure them as well and the first method of making your audit be highly available its DGP clustering so the ditch because drink and you know the tail work was drink is not a new technology in Windows Server and what it does basically is you have two servers that share a same storage and if something happens to one of the servers the failover cluster will detect this and it will fail over all the settings to the second one or depending on how many servers you want in this Belva role and the DGP server will continue to operate in a different node so this is not a new scenario and this is has been going through the years and some the DGP failover on the other hand is the thing that does not require failover clustering to be implemented so this is a question that you can see in some exams what is the difference between these big was drink or ditch buffet over and you need to know that HP fell over is a new tech newer technology that was introduced in Windows Server 2012 and in there you don't need the failover clustering role a feature to configure the DGP failover split scopes on the other hand is another scenario which involves two DGP servers and in this case the two DHCP servers will split the IP addresses between each other for the same scope so if I can show you just real fast if I open the HP console you can just expand it a bit so for example this is going to be my server 0 0 1 that is going to give addresses for my scope and what I want to do with this server I can configure this to give the addresses from 100 to 200 in this example and my second DHCP server can start giving addresses from 2 201 - for example 250 so the server's can work in the same subnet and they can give addresses and if the does the IP addresses from the Scopes are not overlapping everything should be fine and what you can do at you lower the the amount of collisions that are going through when a DGP request is going on if you can configure for example the delay configuration in the Advanced tab of the properties for example on my second server and add some delay so if a client tries to get an IP address on the network this server for example the dish's your one will respond immediately because there is no delay but if this is not reachable my second server will respond after let's say 10 milliseconds delay so this is another way for you to make your DHCP server highly available after giving you some information about the other methods of how you can make your DHCP highly available we are going to take dive into the DGP fell over and this is the latest feature that was introduced in 2012 and to be honest I don't see any difference from 2012 I already made a video but as this is part of the video series of how we can configure the networking in 2016 I think it's necessary for us to repeat this one so what he did be fair over it basically allows us for two servers to communicate between each other and so that they can know when the servers are fully functional and if something happens to one of the servers the other one will take its role to start assigning IP addresses and of course there are multiple modes for DGP for over which we are going to discuss in here so for the first thing I want - um I've already pre-installed on DD ETP roll into my storage server so I'm just going to complete the configuration with the known wizard so if you want to know how to install the ditch below you can check my first video in there I've discussed and Scheldt how to install and configure the ETP now that we have the dhcp roll installed on my end configured and authorized on my storage 0/1 i need to switch to my domain controller where half my scope which I'm going to configure as a DHCP failover scope and I just want you add a huge noting here if you want to configure a DGP fell over this can be done only on to the HP servers so you cannot add more than two DGP servers for DGP forever and you cannot add anything except of from ipv4 so you cannot use these before over for ipv6 at least this is the information that Microsoft are giving up today so I have a single scope and in this code I have a reservation for my PC and I have the lease which is reserved of course and I'm going to use this scope to configure my fail or so in order for me to do this I need to right-click on the ipv4 and configure failover so the first window is the introduction to DHCP fell over and in here if you have multiple scopes you can select all the scopes to make them highly available or you can only choose a single scope in my case I have only one so I'm going to just select all and click Next the second window is asking me to specify the partner server to use for failover and I'm going to add this and you can see that straight away my active directory is detecting that my Storage Server is an authorized dtp server so I'm going to select this option and choose the storage 0 1 so click OK it's going to connect to the server and verify that it is suitable for configuring DGP fell over and we should be okay yes so on the bottom you can see that I have to I have the option to reuse existing federal relationship configured with this server so for example if you configure a single scope with the second DHCP server what you can do is if you decide to make another scope highly available you can go ahead and select the options you use the same settings as the first configuration so you don't need to go through the wizard once again I don't have this option available because this is the initial configuration so I'm going to just leave the default on the next window of the DGP failover wizard you will be presented with the main configurations of your DHCP fell over the first field is the relationship name and the relationship name needs to be unique so for example if you have multiple scopes but you decide to create failover for single scope this name our needs to be configured properly to represent this code so in the future if you decide to make another list for a different scope you need to use this naming convention and configure your next code and again it needs to be unique so in my case what I need to do instead of having the first fqdn then the second text to the end of my servers I will just add vc0 one store 0 1 and then the scope itself so the next time if I have another scope I can just configure the name of the scope or the network of the scope in the last field so this is a good naming convention if you want to configure multiple scopes the next field is the maximum client lead time or NC LT so this is pretty much related to these states which over interval so this state the switch over interval is not enabled by default but if you decide to enable this one and if let's say we have the scenario we're here where we have the two servers and this is 0-1 is my main server and it's giving IP addresses the state switch over interval is the communication between this it monitors the communication between this is 0-1 and storage 0-1 so for example if this is 0 1 is not available not reachable because of let's say network outage storage 0 1 will wait for 60 minutes before actually configuring this 0 1 as down so if you don't have this enabled the server the storage 0 1 server will not automatically assign server down as this is 0 1 at this is year 1 but we will instead need a manual manual configuration from the administrator and when you have this enabled and when your server is detected test down after 60 minutes the maximum Clyde lead time kicks in and then the storage server will wait for another hour before picking up but the the entire scope and start assigning becomes the primary a DHCP server so this is these two are pretty much related to each other it's you need to be familiar with them in order for you to configure your DGP fell over properly so the next field is the mode and you have two options for configuring modes in a DHCP fell over the first one and the default is the load bounce mode and what this basically means is that the two servers will work together to give IP addresses to the clients so in this mode you have at the option to split the ratio on the server so the server's will give the same amount of IP addresses to clients but what you can do is if you decide if you know that your DC is your primary DHCP server and you want this server to host the the main portion of the IP addresses you can change this this percentage for example to 80/20 and it will do it automatically so 80% of the IP addresses will be given to clients from this zero one and only 20% of the addresses will be given by the storage zero one server the next mode is pretty much a disaster recovery situation and this is the hot standby mode in this mode you have a primary server that is actively assigning IP addresses for the scope or the subnet and the second DHCP server assume this role only the primary server becomes unavailable so a DHCP server can simultaneously act as primary for one scope for example and be a secondary for another so if I take example of disco the 10.0.0.0 this is your one can be a primary server for this scope and if I have another one it can be a secondary for another one so it's basically going to be a pretty much an X configuration where you have two scopes that the two separate servers are active on and two scopes that they are standby on so this is a good way for you to configure a proper disaster recovery for your DHCP so within hot standby configuration you have the options you configure what is the role of your partner server and the partner server in my case storage 0 1 so I can configure the partner server to be active or to be on standby and you need to configure the addresses that are reserved for the standby partner the default is 5 percent so in the scenario when my DC 0 1 becomes unavailable and stored the storage server is not able to reach it for example t0 1 is down it will use this 5 percent during the maximum client lead time to assign skills on IP addresses to the clients after the maximum client lead time has passed my storage server will become the active member of the DGP fell over and will the digital one will become distant bye partner so the storage 0/1 will pick up the primary function of the DGP and will start assigning IP addresses to the clients so they will be pretty much no downtime at all so for micro figuration what I'm going to do is I'm going to lower the maximum client lead time to five minutes and to configure an automatic state switch over interval of five minutes and this is not recommended situation because in production if there is a small network outage your HP fell over will constantly go around so I'm doing just to test and confirm that it's fully functional and also I'm going to make my storage zero one my active server and then I'm going to shut down the storage zero one and two wait for ten minutes to pass so I can confirm that this is yo one has become the active member and it has picked up the DGP scope to assign IP addresses for my clients and the last field is the enable message authentication so I'm going to add a secret and this secret needs to be there so that my servers can communicate communicate between each other through a secure Channel so I've added the shared secret and I can click you can verify the settings in here and I will just click finish to configure the DGP available you can see that the process is finished straight away so it's not that it's not it's not taking that much time and now if I switch to my storage zero one server and open my DAT be just let me expand that a bit you will see that under ipv4 I have the scope that we configured on this is 0-1 and all the reservation information and the address pools and everything has has been replicated successfully so to monitor the DGP failover you can go ahead and check it from the properties of the scope you will see that you have another tab in there in there called failover and this is just the read mode only where you can just verify the settings of your two servers and check the state of the of the current server and the partner server and where you can change the settings for the failover are from the ipv4 properties go to the failover tab and you will see that you have maybe more in my situation I have only one configuration and the state of the server is normal again I see some information and if I click Edit I can change the settings of the state switch over change the mode of the DGP failover if I want to and even I can change the shared secret that they shared between the the two servers so to test our DGP configuration what I'm going to do is first I want you remove the reservation for my client computer and you can see that I don't have any reservations at the moment and I'm going to switch to the client computer and let's force a renewal of the IP address renew with IP config renew and let's see if it's able to pick up an IP address from my DHCP server and after successfully now picking up the same ip address as before I can go ahead and once again check my DGP server just refresh the settings and I will see that I have an address lease for this computer so what we are going to do next is I'm going to shut down store 0 1 which is the primary DGP server and we'll wait for 10 minutes to see what is going to happen on my this is 0-1 and if my client is still going to pick up our proper IP address so let's shut down server storage zero one run as administrator shut down st zero and we'll switch to my DC 0 one where if I check the properties and the failover state you will see that the state of this server and has changed to normal from normal to lost contact with partner and the state of the partner server is not available the role on this server is still stand by so if I switch to my client and try to renew the IP configuration let's see if this will be successful so you can see that an error occurred while renewing the IP address of this of this computer and it is unable to contact the DTP server so if I go to yep I would assume that I would need to wait five minutes so that the state of the partner server can be come down so the maximum client lead time will kick in and this server should start giving IP addresses so I'm going to pause here wait for five minutes until the state changes to down and we can continue from there waiting for some time now if I go to the properties and tell what app I'll see that the state of this server is showing partner down so the automatic switch over state has successfully worked and my partner is now down so the role of this server is still standby because it's waiting for the max client lead time to go through but if I switch to my client machine and try to renew the IP settings configuration let's see if this would be successful after playing around a bit with the DGP fell over it seems that what I had to do is I had to first release the IP configuration which is here hi B config release and then IP config renew and what it did it it gave my client machine an IP address from the hot standby ball which you can see starts from 197 so these are the 5% addresses that were reserved by my hot standby server to give while my active server is not online and again I had the real-life scenario where arm my even though my maximum lead time has expired my DHCP server still was showing that it was on standby mode so I looked around I did some configuration and some search and I found that this interval will this state will not change unfortunately I think this is going to be fixed in the future but this is I think known problem that it's only cosmetic just showing that the server is standby and it's not becoming the active member but what I did is I went ahead and configured some additional network cards on my client machine so I've added additional network cards and gave them the automatic configuration on all of them and after switching to my DGP server I noticed some strange behavior so if I open the scope and check the properties under failover I will see that the role of this server is still standby which is again I said it's a strange thing and the failover if I go to the settings I will see that the role of this server is against them by and not the active what I was expecting so what happens in the real life is the server gave IP addresses all the IP addresses that were available from this hot standby and afterwards it did gave the IP addresses from the active partnership that was configured in the favour so even though it's saying that it's hot standby mode is currently on standby it will assign IP addresses from the active member and will give IP addresses up to the client machines that are requesting even in the display statistics fields I found that the server the IP address is available in this server pool are zero and the address is available on the partner pool were 95% it gave the actual IP addresses for the network cards from the partner pool so again after running the full on out-of-state switch over interval and then the maximum lead time indeed it took the ownership of the full scope and it started giving IP addresses for for the clients so you can see that our DHCP failover is fully functional and we even tested the whole standby mode and found some additional things that hopefully in the future there will be fixed so you won't expect anything or you won't be baffled around why your standby server is not becoming the active member when the partner is down so if you liked the video you can always subscribe to our be solutions you can always hit the like button that would be really helpful for helpful for me you can always share the content around the web and show to your colleagues if you don't like the video you can always keep the dislike button leave a comment to tell me what can be improved and we can improve it in the future if you have any questions leave them in the comment section I will try to answer them as soon as possible this was how you can configure DGP fell over on Windows Server 2016 this was Nick from adobe solutions thank you very much for viewing and see you in the next video

Channel: NLB Solutions

Views: 60,522

Rating: 4.9402094 out of 5

Keywords: DHCP Failover, how to configure DHCP failover, DHCP failover Windows Server 2016, implementing windows server 2016 DHCP failover, highly available DHCP, 20-741, 70-741, Dynamic Host Configuration Protocol Failover, DHCP Failover 2017

Id: ijQB5tSBW5Q

Length: 26min 24sec (1584 seconds)

Published: Sun Jun 11 2017