DHCP Failover Modes

Applies To: Windows Server 2012 R2, Windows Server 2012

Two DHCP failover modes are available to use when you create a DHCP failover relationship:

Hot standby mode: This mode provides redundancy for DHCP services.

Load balance mode: This mode allocates DHCP client leases across two servers.

Redundancy for DHCP is also provided in a load-balancing configuration. DHCP clients can renew their leases with the failover partner server if their assigned DHCP server does not respond.

You can switch between hot standby and load balance mode if desired, but you can only use one mode at a time with a single DHCP scope.

For details about settings used for hot standby and load sharing, see DHCP Failover Settings.

For DHCP failover examples using hot standby and load sharing modes, see DHCP Failover Examples.

For information about deploying DHCP failover in hot standby and load sharing mode, see DHCP Failover Architecture.

Hot standby mode

In hot standby mode, two servers operate in a failover relationship where an active server is responsible for leasing IP addresses and configuration information to all clients in a scope or subnet. The partner server assumes a standby role, with responsibility to issue leases to DHCP clients only if the active server becomes unavailable. Hot standby mode is ideal for scenarios where the failover partner is only intended to be used temporarily when the active server is unavailable.

A server is active or standby in the context of a failover relationship. For instance, a server that has the role of active for a given relationship could be a standby server for another relationship. By default, the server that is used to create the failover relationship is the active server, but this is not required.

When you choose hot standby, you must also configure the percentage of IP addresses on the active server that are reserved for use on the standby server in the event that the active server does not respond. By default, this reserve percentage is 5%.

The reserve percentage is used for new DHCP leases. If a DHCP client attempts to renew a DHCP lease with the standby server that is unable to contact the active server (COMMUNICATIONS INTERRUPTED or PARTNER DOWN state exists), the same IP address that was previously assigned to the DHCP client will be renewed. In this situation, a temporary lease is granted for the maximum client lead time (MCLT) duration, not the full scope lease time.

In a COMMUNICATIONS INTERRUPTED or PARTNER DOWN state, if the standby server issues all its available reserve percentage leases to new DHCP clients before the MCLT expires, it will refuse to issue new DHCP leases, but it will continue to renew existing leases. After the MCLT has expired, the standby server will be permitted to use the entire available IP address pool for new DHCP leases, provided the server is in PARTNER DOWN state. If the server is still in COMMUNICATIONS INTERRUPTED state, it will not use the entire available IP address pool for new DHCP leases.

For more information about DHCP failover settings and states, see DHCP Failover Settings.

Load balance mode

Load balance mode is the default mode of deployment. In this mode, two DHCP servers simultaneously serve IP addresses and options to clients on a given subnet. DHCP client requests are load balanced and shared between the two DHCP servers. The default load balancing ratio between the two servers is 50:50, but this can be customized to any ratio from 0 to 100%.

The load-balancing mechanism is defined in RFC 3074, in which a hash is computed from the MAC address contained in each DHCP client request. A range of hash values (also called the hash bucket) is assigned to each DHCP server based on the load balancing percentages that are configured. Servers determine if they are designated to respond to the client based on their assigned hash bucket.

In load balancing mode, when a DHCP server loses contact with its failover partner it will begin granting leases to all DHCP clients. If it receives a lease renewal request from a DHCP client that is assigned to its failover partner, it will temporarily renew the same IP address lease for the duration of the MCLT. If it receives a request from a client that was not previously assigned a lease, it will grant a new lease from its free IP address pool until this is exhausted, and then it will begin using the free IP address pool of its failover partner. If the DHCP server enters a partner down state, it will wait for the MCLT duration and then assume responsibility for 100% of the IP address pool.

Windows server in standby

This forum has migrated to Microsoft Q&A. Visit Microsoft Q&A to post new questions.

Answered by:

Question

I have 2 Windows Server 2008R2 with 8GB memory running on Vmware ESXi 5.0. I did a reboot on them last night and now 10 hours later one of the servers has no memory free. The other server has over 3 GB free.
Both servers are Fileserver for our Citrix enviroment.
Server 1:

Why is they different?
I would like them to have more free memory.
Why do one of the servers take everything to standby?

Answers

Standby memory is memory that is currently not in use, for example if you would have a process that was using 1GB of memory and you close the program. It can occur that the 1GB does not become free memory, but stays in memory as standby memory waiting to be recalled at a later moment.

Have a look at this similar thread and the links listed in the answer for more information about memory in Server 2008:

All replies

There is no way to tell from these pictures, possible causes are:

• Higher load on the server with higher memory requirement
• Application with memory leak
• Driver with memory leak (not likely with ESX)

What would be more interesting to see is to have a look at the performance tab of both VMs and check how much memory ESX actually has assigned to those two machines. Based on that you can usually tell if it is actually using all the memory Windows thinks it is using.

Thanks for the answer.
We have Transparent Page Sharing on our VMware environment, so what the performance tab say is different from what windows say.

But what I want to learn here, is what is the difference between free and standby memory?
Why does it sometimes take everything in standby and other times just have free memory?

Standby memory is memory that is currently not in use, for example if you would have a process that was using 1GB of memory and you close the program. It can occur that the 1GB does not become free memory, but stays in memory as standby memory waiting to be recalled at a later moment.

Have a look at this similar thread and the links listed in the answer for more information about memory in Server 2008:

Windows server in standby

Вопрос

Hi all, I have an existing DHCP failover relationship between 2 servers. The relationship is configured in Hot Standby mode. There are approximately 35 scopes. Currently DHCP2.mydomain.com is the Active server, and DHCP1.mydomain.com is the Standby. Management would like me to switch it so that DHCP1.mydomain.com is the active and DHCP2.mydomain.com is the standby server.

Output of Get-DHCPserverv4Failover on DHCP2.mydomain.com below.

Name : dhcp2.mydomain.com-dhcp1.mydomain.com
PartnerServer : dhcp1.mydomain.com
Mode : HotStandby
LoadBalancePercent :
ServerRole : Active
ReservePercent : 5
MaxClientLeadTime : 01:00:00
StateSwitchInterval :
State : Normal
ScopeId : <10.128.4.0, 10.128.7.0, 10.128.155.0, 10.128.153.0. >
AutoStateTransition : False
EnableAuth : True

Output of Get-DhcpServerv4Failover when run on DHCP1.mydomain.com

Name : dhcp2.mydomain.com-dhcp1.mydomain.com
PartnerServer : dhcp2.mydomain.com
Mode : HotStandby
LoadBalancePercent :
ServerRole : Standby
ReservePercent : 5
MaxClientLeadTime : 01:00:00
StateSwitchInterval :
State : Normal
ScopeId : <10.128.4.0, 10.128.7.0, 10.128.155.0, 10.128.153.0. >
AutoStateTransition : False
EnableAuth : True

I am thinking I can use the set-dhcpserverv4failover command to make the change I am desiring, but not quite sure of the syntax. Can anyone provide the correct syntax to swap the standby and active servers in an existing DHCP failover relationship? Thanks.

Все ответы

>I am thinking I can use the set-dhcpserverv4failover command to make the change I am desiring, but not quite sure of the syntax.

With «set-dhcpserverv4failover command», the Set-dhcLoadBalancePercent, MaxClientLeadTime, StateSwitchInterval, AutoStateTransition, and SharedSecret parameters can be specified. While we can’t use this command to change the active and hot standby mode:

And after researching, seems no command to change the standby and active mode in an existing DHCP failover relationship neither via Commandlet nor DHCP GUI. We need to re-configure the DHCP failover to change it.

>However, unless I’m mistaken, when I delete the failover relationship, all the leases are doing to be removed from DHCP1 and then I would need to export them all from DHCP2 to an xml file and import into DHCP1. I would rather not go through that process if possible.

I tested in my lab, when we re-configure the DHCP failover relationship, the DHCP lease won’t be affected. Clients won’t discover the changes in DHCP server.

On the original DHCP failover, right click the scope>Deconfigure Failover, refresh the master and the partner server, then configure failover again, this time, choose the partner DHCP server to be active mode.

Windows server in standby

Might also try over here.

Regards, Dave Patrick .
Microsoft Certified Professional
Microsoft MVP [Windows Server] Datacenter Management

Disclaimer: This posting is provided «AS IS» with no warranties or guarantees, and confers no rights.

Standby memory are Pages of physical ram not actively being used. These are still left in physical ram but will be repurposed first by the memory manager (either returned to the active list or zeroed out and reused) if something needs physical ram for active pages.

Please reference the article below for further understanding:

Where has all my Physical RAM gone?

Please remember to mark the replies as answers if they help.
If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

The Standby list, which is shown in blue, contains pages that have been removed from process working sets but are still linked to their respective working sets. As such, Standby list is essentially a cache. However, memory pages in the Standby list are prioritized in a range of 0-7, with 7 being the highest. Essentially, a page related to a high-priority process will receive a high-priority level in the Standby list.

For example, processes that are Shareable will be a high priority and pages associated with these Shareable processes will have the highest priority in the Standby list.

Now, if a process needs a page that is associated with the process and that page is now in the Standby list, the memory manager immediately returns the page to that process’ working set. However, all pages on the Standby list are available for memory allocation requests from any process. When a process requests additional memory and there is not enough memory in the Free list, the memory manager checks the page’s priority and will take a page with a low priority from the Standby list, initialize it, and allocate it to that process.

And you could use RAMMap to free up some memory,here is the download link:

Best Regards,
Cartman
Please remember to mark the replies as an answers if they help and unmark them if they provide no help.
If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

Standby memory are Pages of physical ram not actively being used. These are still left in physical ram but will be repurposed first by the memory manager (either returned to the active list or zeroed out and reused) if something needs physical ram for active pages.

Please reference the article below for further understanding:

Where has all my Physical RAM gone?

Please remember to mark the replies as answers if they help.
If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

The Standby list, which is shown in blue, contains pages that have been removed from process working sets but are still linked to their respective working sets. As such, Standby list is essentially a cache. However, memory pages in the Standby list are prioritized in a range of 0-7, with 7 being the highest. Essentially, a page related to a high-priority process will receive a high-priority level in the Standby list.

For example, processes that are Shareable will be a high priority and pages associated with these Shareable processes will have the highest priority in the Standby list.

Now, if a process needs a page that is associated with the process and that page is now in the Standby list, the memory manager immediately returns the page to that process’ working set. However, all pages on the Standby list are available for memory allocation requests from any process. When a process requests additional memory and there is not enough memory in the Free list, the memory manager checks the page’s priority and will take a page with a low priority from the Standby list, initialize it, and allocate it to that process.

And you could use RAMMap to free up some memory,here is the download link: