by Michael Bose

VMware vSphere HA and DRS Compared and Explained

A VMware hypervisor allows you to run virtual machines on a single server. You can run multiple virtual machines on a standalone ESXi host and deploy multiple hosts to run more virtual machines. If you have multiple ESXi hosts connected via the network, you can migrate virtual machines from one horde to another. sometimes using multiple hosts connected via the network to run VMs is not adequate to meet clientele needs. For example, in cases where one server fails, all VMs residing on that server will besides fail. besides, VM workloads on ESXi hosts can be imbalanced, and migrating VMs manually between hosts is routine. To deal with these issues, VMware provides clustering features such as VMware High Availability ( HA ) and Distributed Resource Scheduler ( DRS ). Using vSphere clustering allows you to reduce VM downtime and consume hardware resources rationally. This web log stake covers VMware HA and DRS adenine well as the use cases for each bunch sport. Before we start

VMware HA helps you to reduce downtime in font an ESXi host fails as a leave of hardware failure. however, there can be other types of failures such as software failure inside a VM or data passing caused by human error or malware. Best practices for running virtual machines in VMware vSphere recommend regular VMware backup of VMs even if VMware DRS and HA are enabled. Watch this webinar about VMware VM data protection with NAKIVO Backup & Replication to learn more about VMware vSphere VM accompaniment features .

What Is a vSphere Cluster?

A vSphere bunch is a laid of connected ESXi hosts that contribution hardware resources such as central processing unit, memory, and storage. VMware vSphere clusters are centrally managed in vCenter. A bunch ’ randomness resources are aggregated into a resource consortium, so when you add a master of ceremonies to a bunch, the master of ceremonies ’ mho resources become character of the resources of the whole cluster. ESXi hosts that are members of the cluster are besides called bunch nodes. There are two types of vSphere clusters : vSphere High Availability and Distributed Resource Scheduler ( VMware HA and DRS ).

VMware cluster requirements

To deploy VMware HA and DRS, a dress of bunch requirements should be met :

  • Two or more ESXi hosts with an identical configuration (processors of the same family, ESXi version and patch level, etc.) must be used. For instance, you can use two servers with Intel processors of the same family (or AMD processors) and ESXi 7.0 Update 3 installed on the servers. It is recommended that you use at least three hosts for better protection and performance.
  • High-speed network connections for the management network, storage network, and vMotion network. Redundant network connections are required.
  • A shared datastore that is accessible for all ESXi hosts within a cluster. Storage Area Network (SAN), Network Attached Storage (NAS), and VMware vSAN can be used as a shared datastore. NFS and iSCSI protocols are supported to access data on a shared datastore. VM files must be stored on a shared datastore.
  • VMware vCenter Server that is compatible with the ESXi version installed on the hosts.

Unlike a Hyper-V Failover Cluster, no quorum is required and you don ’ t need to use complex network names.

What is VMware HA in vSphere?

VMware vSphere High Availability ( HA ) is a bunch feature that is designed to restart a virtual machine ( VM ) mechanically in case of failure. VMware vSphere High Availability allows organizations to ensure high handiness for VMs and applications running on the VMs in a vSphere bunch ( independent of running applications ). VMware HA can provide protection against the failure of an ESXi horde – the fail VM is restarted on a goodly host. As a result, you can importantly reduce downtime.

Requirements for vSphere HA

Requirements for vSphere HA must be taken into account together with the general vSphere bunch requirements. In arrange to set up VMware vSphere High Availability, you must have :

  • A VMware vSphere Standard license
  • Minimum 4 GB of RAM on each host
  • A pingable gateway

How Does vSphere HA Work?

VMware vSphere High Availability checks ESXi hosts to detect a host bankruptcy. If a server failure is detected ( VMs running on that host are besides failed ), then the failed VMs are migrated to healthy ESXi hosts within the bunch. After migration, the VMs are registered on the new hosts, and then these VMs are started. VM files ( VMX, VMDK, and other files ) are located on the lapp resource, which is a shared datastore, after migration. VM files are not migrated. only the CPU, memory, and network components utilized by the fail VMs are provided by the raw ESXi host after migration. The downtime is equal to the time needed to restart a VM on another server. however, keep in mind that there ’ second besides the prison term needed for the operate arrangement to boot and for loading the needed applications on a VM. VMware HA is a solution that works on the VM layer and can be besides used if applications don ’ t have native senior high school handiness features. VMware vSphere High Availability doesn ’ t depend on the guest operate on organization installed on the VM. The work flow of a vSphere HA bunch is illustrated in the diagram below. There is a cluster with three ESXi hosts in this example. VMs are running on all hosts. The connections of the VMs and their files are illustrated with dot lines. 1. The normal operation of a bunch. All VMs are running on their native hosts. The normal operation of a vSphere HA cluster The normal operation of a vSphere HA cluster 2. ESXi host 1 fails. VMs residing on ESXi host 1 ( VM1 and VM2 ) are failed ( these VMs are powered off ). A vSphere HA bunch initiates VM restarts on other healthy ESXi hosts. One ESXi host fails and vSphere HA initiates VM failover One ESXi host fails and vSphere HA initiates VM failover 3. VMs have been migrated and restarted on healthy hosts. VM1 has been migrated to ESXi host 2, and VM2 has been migrated to ESXi master of ceremonies 3. VM files are located in the lapp place on the shared storage that is connected to all ESXi hosts of the vSphere cluster. VM failover has finished and all VMs are running in the vSphere HA cluster VM failover has finished and all VMs are running in the vSphere HA cluster

HA master and subordinates

After vSphere senior high school Availability is enabled in the bunch, one ESXi host is selected as the HA dominate. other ESXi hosts are subordinates ( slave subordinate hosts ). A master monitors the state of the subordinates to detect host failure in time and initiate the resume of fail VMs. The maestro server besides monitors the power state of matter of virtual machines on cluster nodes. If a VM failure is detected, the master initiates a VM restart ( the optimum horde is selected by the master before restarting the fail VM ). The HA chief sends data about the HA cluster ’ s health to vCenter. VMware vCenter manages the bunch by using the interface provided by the HA passkey server. The master can run VMs just as early hosts within the cluster. If a victor host fails, another headmaster horde is selected. The host that is connected to the highest number of datastores has the advantage in the election of the primary ESXi host. Hosts that are not in the alimony manner participate in the election of the primary horde. dependent hosts can run virtual machines, monitor the states of VMs, and report updated information about VM states to the HA master host. Fault Domain Manager ( FDM ) is the appoint of the agent used to monitor the handiness of physical servers. The FDM agent works on each ESXi master of ceremonies within an HA bunch.

Host Failure Types

There are three types of ESXi host failures : Failure. An ESXi host has stopped working for some reason. Isolation. An ESXi server and the VMs on this host continue to work, but the host is isolated from early hosts in the cluster due to network issues. Partition. The network connectivity with the primary master of ceremonies is lost.

How failures are detected

Heartbeats are exchanged to detect failures in a vSphere HA bunch. The primary host monitors the country of secondary hosts by receiving heartbeats from the secondary hosts every second. The chief host sends ICMP pings to the secondary host and waits for replies. If the primary host can not communicate directly with the agent of the secondary host, the secondary master of ceremonies can be healthy or failed but inaccessible via the network. If the primary host doesn ’ metric ton meet heartbeats, then the primary coil host checks the leery host by means of Datastore Heartbeating. During convention operation, each host within an HA cluster exchanges heartbeats with the shared datastore. The primary ESXi host checks if the datastore heartbeats have been exchanged with the leery host in summation to sending pings to that host. If there is no datastore blink of an eye change with the leery host, and that server doesn ’ deoxythymidine monophosphate commit ICMP requests, then the host is designated as a fail host. Note: A particular .vSphere-HA directory is created at the root of a shared datastore for heartbeating and identifying a list of protect VMs. note that vSAN datastores can not be used for datastore heartbeating. If the elementary server can not connect with the agent of the secondary host, but the secondary host exchanges heartbeats with the shared datastore, then the elementary host marks the fishy horde as a network detached host. If the chief host determines that the secondary host is running in an disjunct network section, the primary horde continues to monitor the VMs on that isolated host. If the VMs on the isolated host are powered off, then the primary master of ceremonies initiates the restart of these VMs on another ESXi host. You can configure the vSphere HA cluster ’ s response to an ESXi server becoming net isolated. Monitoring individual VMs. VMware vSphere High Availability has a mechanism for monitoring person VMs and detecting if a especial VM has failed. VMware Tools installed on a node function organization ( OS ) are used to determine the virtual machine state of matter. VMware Tools send guest OS heartbeats to the ESXi host. Heartbeats and input/output ( I/O ) activeness generated by VMware Tools are monitored by the VM monitoring service. If the elementary ESXi host in the HA cluster detects that VMware Tools on the protected VM are not responding and that there is no I/O natural process, the host initiates a VM resume. Monitoring VM I/O activeness allows an HA cluster to avoid unnecessary VM resets if VMware Tools don ’ thymine send heartbeats for some reason but the VM is running. You can set monitor sensitivity to configure the time period after which a VM must be restarted if guest OS heartbeats generated by VMware Tools are not received by the ESXi host. VMware vSphere HA restarts the VM on the lapp ESXi host in case of a single VM failure. VMware Tools heartbeats are sent to hostd on the hypervisor charge ( ESXi ), not by using the network stack. then the ESXi host sends the pick up information to vCenter. VMware Tools heartbeats can be received by an ESXi host if a VM is disconnected from a network and flush if there is no virtual network adapter connected to the VM. VM and application monitoring. You can use SDK from a third-party seller to monitor whether a specific application installed on a VM has failed. The alternative option is to use an application that already supports VMware Application Monitoring. Application heartbeats are used for application monitoring on VMware VMs running in a vSphere HA bunch.

Key Parameters for HA Cluster Configuration

Before you start to configure an HA cluster, you need to define some key parameters. Isolation response is a argument that defines how an ESXi host acts when it does not receive heartbeat signals. The options are Leave powered on, Power off ( default ), and Shutdown. Reservation is a argument that is calculated based on the maximal characteristics of the most resource-hungry VM within a bunch. This parameter is used for estimating Failover Capacity. An HA cluster creates reservation slots by using the value of the Reservation parameter. Failover Capacity. This parameter is measured in integer and defines the maximum act of servers that can be failed in the bunch without a negative shock on workloads ( the cluster and all VMs can calm operate after the failure of this number of ESXi hosts ). The number of host failures allowed. This parameter is defined by a system administrator to set how many hosts can fail to continue the operation of the bunch. Failover Capacity is taken into account when setting the value for this parameter. Admission Control is the parameter used to ensure that there are sufficient resources reserved to recover VMs after an ESXi host fails. This parameter is set by an administrator and defines the behavior of the VMs if there are not enough free slots to start VMs after ESXi host failures. Admission Control defines the failover capability, that is, the percentage of resource abasement that can be tolerated in a vSphere HA bunch after failover. Restart Priority is set by an administrator to define the succession for starting VMs after the failover of a cluster node. Administrators can configure vSphere HA to start critical virtual machines first and then start early VMs.

Failover capacity and host failure

Let ’ s look at two cases, each with three ESXi hosts but with different failover capacitance values. In the first gear subject, the HA bunch can work after the failure of one ESXi host ( see the leave side of the image below ). In the second subject, the HA bunch can tolerate the failure of two ESXi hosts ( see the correct side of the image ). 1. Each ESXi horde has 4 slots. There are 6 VMs in the bunch. If one ESXi horde fails ( the third master of ceremonies, for example ), then the three VMs ( VM4, VM5, and VM6 ) can migrate to the early two ESXi hosts. In my exemplar, these three VMs are migrating to the irregular ESXi host. If one more ESXi host fails, then there will be no detached slots to migrate and run other VMs. 2. Each ESXi host has 4 slots. 4 VMs are running in the VMware vSphere HA cluster. In this case, there are adequate slots to run all VMs within a cluster if two ESXi hosts fail. What is VMware HA failover capacity What is VMware HA failover capacity In order to calculate Failover Capacity, do the follow : From the number of all nodes in the bunch subtract the proportion of the issue of VMs in the cluster to the number of slots on one node. If you have a non-integer ( a number that is not wholly ) as a result, round the number to the nearest last integer. Let ’ s account Failover Capacity for the two examples. Example 1: 3–6/4=1.5 polish 1.5 to 1. All VMs in an HA bunch can survive if 1 ESXi horde fails. Example 2: 3–4/4=2 There is no indigence to round down, as 2 is an integer. All VMs can continue operating if 2 ESXi hosts fail.

Admission Control

As mentioned above, entree control is the parameter needed to ensure that there are enough resources to run VMs after a host failure in the cluster. You can besides define the Admission Control State parameter for more convenience. Admission Control State is calculated as the proportion of Failover Capacity to Number of Host Failures allowed ( NHF ). If Failover Capacity is higher than NHF, then the HA bunch is configured by rights. otherwise, you need to set Admission Control manually. There are two options available : 1. Do not power on virtual machines if they violate handiness constraints ( do not world power on VMs if there are not enough hardware resources ). 2. Allow virtual machines to be started even if they violate handiness ( start VMs despite the lack of hardware resources ). Choose the choice that fits your vSphere High Availability cluster consumption case best. If your objective is the dependability of the HA cluster, select the beginning option ( Do not power on VMs ). If the most important thing for you is running all VMs, then select the second option ( Allow VMs to be started ). Be mindful that in the moment case the demeanor of the bunch can be unpredictable. In the worst-case scenario, the HA cluster can become useless.

VM overrides

VM overrides ( or HA overrides in the case of an HA cluster ) the choice that allows you to disable HA for a particular VM ladder in the HA cluster. You can configure your vSphere HA cluster at a more chondritic horizontal surface with this option on the cluster level.

Fault Tolerance

VMware provides a have for a vSphere HA bunch that allows you to achieve zero downtime in case of an ESXi server failure. This feature of speech is called Fault Tolerance. While the standard configuration of vSphere High Availability requires a VM restart in case of failure, Fault Tolerance allows VMs to continue running if the chief ESXi host on which the VMs are registered fails. Fault Tolerance can be used for mission-critical VMs running critical applications. There is the overhead to achieve zero downtime for the highest level of business continuity because there are two running instances of a VM protected with Fault Tolerance. The second haunt VM is running on the second ESXi host, and all changes to the original VM ( CPU, RAM, network state ) are replicated from the initial ESXi host to the secondary ESXi master of ceremonies. The protect VM is called the chief VM and the extra VM is called the secondary VM. The primary coil and secondary VMs must reside on unlike ESXi hosts to ensure protection against ESXi server failure. The two VMs ( primary VM and secondary VM ) are running simultaneously and consume CPU, RAM, and network resources on both ESXi hosts ( therefore a VM protected with the Fault Tolerance sport consumes twice equally many resources in the vSphere HA cluster ). These VMs are endlessly synchronized in real-time. Users can work alone with the primary ( master ) VM and the secondary ( ghost ) VM is invisible to them. How Fault Tolerance works in the two-node vSphere HA cluster How Fault Tolerance works in the two-node vSphere HA cluster

If the first ESXi host fails ( the host on which the basal VM is residing ), then the workloads are migrated to the secondary VM ( that is the VM knockoff or the ghost VM ) running on the irregular ESXi host. The secondary VM becomes active and accessible in a moment. Users can notice a flimsy network latency during the moment of the guileless failover. There is no service pause or data loss during failover. After the failover was done successfully, a raw ghost VM is created on the option goodly ESXi host to provide redundancy and continue VM auspices against ESXi host failure. Seamless failover of a VM in a two-node vSphere HA cluster with Fault Tolerance Seamless failover of a VM in a two-node vSphere HA cluster with Fault Tolerance fault Tolerance avoids split-brain scenarios ( when two active copies of a protect VM run simultaneously ) due to the file locking mechanism on shared storage for failover coordination. however, Fault Tolerance doesn ’ t protect against software failures inside a VM ( such as guest OS failure or failure of particular applications ). If a primary coil VM fails, the secondary VM besides fails. Requirements for Fault Tolerance

  • A vSphere HA cluster with a minimum of two ESXi hosts.
  • vMotion and FT logging.
  • A compatible CPU that supports hardware-assisted MMU virtualization.

Using a dedicated Fault Tolerance network in the vSphere HA cluster is recommended. A license for Fault Tolerance

  • ESXi hosts must be licensed to use Fault Tolerance.
  • vSphere Standard and Enterprise support up to 2 vCPUs for a single VM.
  • vSphere Enterprise Plus allows you to use up to 8 vCPUs per VM.

Fault Tolerance limitations There are some limitations to using VMware Fault Tolerance in vSphere. VMware vSphere features that are antagonistic with FT :

  • VM snapshots. A protected VM must not have snapshots.
  • Linked clones
  • VMware vVol datastores

Devices that are not supported:

  • Raw device mapping devices
  • Physical CD-ROM and other devices of a server that are connected to a VM as virtual devices
  • Sound devices and USB devices
  • VMDK virtual disks which size is more than 2 TB
  • Video devices with 3D graphics
  • Parallel and serial ports
  • Hot-plug devices
  • NIC (network interface controller) pass-through
  • Storage vMotion (must be temporarily disabled to migrate VM files to another storage)

What is DRS in VMware vSphere?

Distributed Resource Scheduler ( DRS ) is a VMware vSphere clustering feature that allows you to load remainder VMs linear in the cluster. DRS checks the VM burden and a load of ESXi servers within a vSphere bunch. If DRS detects that there is an clog host or VM, DRS migrates the VM to an ESXi host with enough exempt hardware resources to ensure the quality of service ( QoS ). DRS can select the optimum ESXi host for a VM when you create a new VM in the bunch. VMware DRS allows you to run VMs in a poise bunch and avoid overload and situations when there are not enough hardware resources for virtual machines and applications running on VMs for normal operation ( there must be enough resources in the whole bunch in this case ).

DRS Requirements

The requirements for DRS, along with the general requirements for a vSphere bunch, include :

  • vSphere Enterprise or vSphere Enterprise Plus license
  • A CPU with Enhanced vMotion Compatibility for VM live migration with vMotion
  • A dedicated vMotion network

A configured VMware vMotion is required to operate a DRS bunch, unlike an HA bunch, where vMotion is required only if using Fault Tolerance. besides, the vSphere license required for VMware DRS is higher than the license for using vSphere high Availability.

The role of vMotion

Migrate VMs from one ESXi host to another with vMotion, which we mentioned when explaining how Fault Tolerance works. With VMware vMotion, VM migration ( CPU, memory, network state of matter ) occurs without interrupting running VMs ( there is no downtime ). VMware vMotion is the key feature of speech for the proper function of DRS. Let ’ s expect at the chief steps of vMotion operation : 1. vMotion creates a shadow VM on the destination ESXi horde. The finish ESXi server pre-allocates enough resources for the VM to be migrated. The VM is put in the intercede country and the VM shape can not be changed during migration. Preparing for VM migration with vMotion Preparing for VM migration with vMotion 2. The precopy process. Each VM memory page is copied from the source to the destination by using a vMotion network. 3. The future fall of copying memory pages from the source to the finish is performed, because memory pages are being changed during the VM operation. This is an iterative process that is performed until no changed memory pages remain. The change memory pages are called dirty pages. It takes more clock for VM migration with vMotion if memory-intensive operations are performed on a VM because more memory pages are changed. How vMotion copies VM memory How vMotion copies VM memory 4. The VM is stopped on the source ESXi host and resumed on the address host. Insignificant network latency can be noticed inside the migrate VM for about a second at this moment. Finishing VM migration with vMotion in vSphere Finishing VM migration with vMotion in vSphere

The Working Principle of DRS in VMware

VMware DRS checks the workloads from the CPU and RAM perspectives to determine the vSphere cluster symmetry every 5 minutes, which is the nonpayment interval. VMware DRS checks all of the resources in the resource pool of the bunch, including the resources consumed by VMs and the resources of each ESXi host within the bunch that can be provided to run VMs. Resource checks are performed according to the configure policies. The demands of VMs are besides taken into history ( the hardware resources that the VM needs to run at the moment of checking ). The formula is used for VM demand calculation for memory : VM memory demand = Function(Active memory used, Swapped, Shared) + 25% (idle consumed memory) CPU necessitate is calculated based on the number of processor resources that are presently consumed by a VM. VM CPU maximal values and VM CPU average values collected during the last hindrance help the DRS to determine the course of resource custom for a detail VM. If vSphere DRS detects an asymmetry in the bunch and that some ESXi hosts are overloaded, then the DRS initiates live migration of VMs running on the clog server to a host with free resources. Let ’ s expect at how vSphere DRS works in VMware using an exercise with diagrams. In the diagram below, you can see a DRS cluster with 3 ESXi hosts. All the hosts are connected to shared memory, where VM files are located. The first host is very load, the second server has release CPU and memory resources, and the third host is heavily loaded. Some VMs on the first ( VM1 ) and one-third ( VM4, VM5 ) ESXi hosts are consuming about all the provisioned CPU and memory resources. In this case, the performance of these VMs can degrade. DRS in VMware vSphere – a cluster that is not balanced DRS in VMware vSphere – a cluster that is not balanced VMware DRS determines that the rational action is to migrate the heavily loaded VM2 from the clog ESXi horde 1 to ESXi host 2, which has adequate free resources and migrate VM4 from ESXi host 3 to ESXi host 2. If the DRS is configured to work in the automatic rifle modality, then running VMs are migrated with vMotion ( this legal action is illustrated with green arrows in the prototype below ). VM files including virtual disks ( VMDK ), configuration files ( VMX ), and other files are located in the same stead on the shared storage during VM migration and after VM migration ( connections of VMs and their files are illustrated with dotted lines on the picture ). once the selected VMs have been migrated, the DRS cluster become balanced. There are free resources on each ESXi horde within the bunch to run VMs effectively and ensure high performance. The VMware DRS cluster is balanced after VM migration The VMware DRS cluster is balanced after VM migration The site can change due to mismatched VM workloads, and the bunch can become unbalanced again. In this case, DRS will check the resources consumed and release resources in the cluster to initiate VM migration again.

Key Parameters for vSphere DRS Configuration

VMware vSphere DRS is a highly customizable clustering sport that allows you to use DRS with higher efficiency in different situations. Let ’ s look at the independent parameters that affect the demeanor of DRS in a vSphere bunch.

VMware DRS automation levels

When DRS detects that a vSphere bunch is imbalanced, DRS provides recommendations for VM placement and migration with vMotion. The recommendation can be implemented by using one of three automation levels : Fully automated. Initial VM placement and vMotion recommendations are applied mechanically by DRS ( exploiter intervention is not required ). Partially automated. Recommendations for the initial placement of new VMs are the entirely ones applied mechanically. other recommendations can be initiated and applied manually or ignored. Manual. DRS provides recommendations for initial VM placement and VM migration but drug user interaction is required to apply these recommendations. You can besides ignore the recommendations provided by DRS.

DRS aggression levels (migration thresholds)

DRS aggression levels or migration thresholds is the option to control the utmost imbalance level that is acceptable for a DRS bunch. There are five doorsill values from 1, the most conservative, to 5, the most aggressive. The aggressive set initiates VM migration tied if the benefit of VM placement is little. The conservative context doesn ’ t initiate VM migration flush if meaning benefits can be achieved after VM migration. Level 3, the middle aggression level, is selected by default and this is the commend fix.

Affinity Rules in VMware DRS

Affinity and anti-affinity rules are utilitarian when you need to place specific VMs on specific ESXi hosts. For model, you may need to run some VMs together on one ESXi host within a bunch, or frailty versa ( you need two or more VMs to be placed merely on different ESXi hosts, and VMs must not be placed on one host ). function cases can include :

  • Virtual domain controller VMs (a primary domain controller and additional domain controller) on different hosts to avoid the failure of both VMs if one host fails. These VMs must not run together on a single ESXi host in this case.
  • VMs running software that is licensed to run on the appropriate hardware and cannot run on other physical computers due to licensing limitations (for example, Oracle Database).

affinity rules are divided into :

  • VM-VM affinity rules (for individual VMs)
  • VM-host affinity rules (relationship between groups of hosts and groups of VMs)

VM host rules can be discriminatory ( VMs should… ) and mandate ( VM must… ). mandate rules continue to work tied if DRS is disabled, which doesn ’ metric ton allow you to migrate the allow VMs with vMotion manually. This rationale is used to avoid trespass of the rule applied to VMs running on ESXi hosts if vCenter is temporarily unavailable or failed. There are four options for DRS affinity rules : Keep virtual machines together. The selected VMs must run together on a single ESXi host ( if VM migration is needed, all these VMs must be migrated together ). This rule can be used when you want to localize network traffic between the selected VMs ( to avoid network overloading between ESXi hosts if VMs beget significant network traffic ). Another use case is running a complex application that uses components ( that depend on each other ) installed on multiple VMs or running a vApp. This could include, for exemplar, a database server and an application waiter. Separate virtual machines. The selected VMs must not run on a unmarried ESXi host. This option is used for high handiness purposes. Virtual machines to hosts. VMs added to a VM group must run on the specified ESXi host or server group. You need to configure DRS groups ( VM/Host groups ). A DRS group contains multiple VMs or ESXi hosts. Virtual machines to virtual machines. This rule can be selected to tie VMs to VMs when you want to power on one VM group and then power another ( dependent ) VM group. This option is used when VMware HA and DRS are configured together in the cluster. If there is a rule conflict, then the older principle takes precedence.

VM Override for VMware DRS

alike to the use of VM override in a vSphere HA bunch, VM overrides are used for more chondritic configurations of DRS in VMware vSphere and allow you to override ball-shaped settings set at the DRS bunch level and define specific settings for an person VM. other VMs of the bunch are not affected when VM override is applied for a specific VM.

Predictive DRS

The main concept of Predictive DRS is to collect information about VM placement and then, based on the previously collected information, predict when and where high gear resource custom will occur. Using this information, Predictive DRS can move VMs between hosts for better load balancing before an ESXi server is overloaded and VMs lack resources. This sport can be utilitarian when there are time-based need changes for VMs in a bunch. predictive DRS is disabled by default. VMware vRealize Operations Manager is required to use Power DRS. Predictive DRS in VMware vSphere Predictive DRS in VMware vSphere

Distributed Power Manager

Distributed Power Manager ( DPM ) is a sport used to migrate VMs if there are enough barren resources in a bunch to shut down an ESXi master of ceremonies ( put a host into the standby mode ) and run VMs on the remaining ESXi hosts within the bunch ( remaining hosts must provide enough resources to run the need VMs ). Distributed Power Manager has detected free resources in the vSphere DRS cluster Distributed Power Manager has detected free resources in the vSphere DRS cluster When more resources are needed in a bunch to run VMs, DPM initiates a server that was shut down to wake up and operate in normal modality. One of the corroborate exponent management protocols is used to ability on a server via the network. These protocols are intelligent Platform Management Interface ( IPMI ), Hewlett-Packard Integrated Lights-Out ( international labor organization ), or Wake-On-LAN ( WOL ). then DRS migrates some VMs to this waiter to distribute workloads and balance a cluster. By default, Distributed Power Management is disabled. DPM recommendations can be applied automatically or manually. Distributed Power Manager migrates all VMs from an ESXi host to shut down the host Distributed Power Manager migrates all VMs from an ESXi host to shut down the host

Storage DRS

While DRS migrates VMs based on CPU and RAM calculation resources, Storage DRS migrates virtual machine files from one datastore to another based on datastore custom, for exemplar, free harrow outer space. Affinity and anti-affinity rules allow you to configure whether storage DRS must store a VM ’ s virtual harrow files together on the same datastore. For example, you can configure the anti-affinity rule to store the VMDK files of a VM that performs I/O intensive operations on different datastores. You do this to avoid performance debasing of the VM and initial VM datastore ( I/O magnetic disk workloads will be distributed across multiple datastores when using the anti-affinity rule ). memory DRS is useful when using VMs with sparse provision disks in event of overprovisioning. storage DRS helps avoid situations when the size of thin disks grows, and as a result, there is no free outer space on a datastore. Lack of absolve distance causes the VMs storing virtual disks on that datastore to fail. virtual machine phonograph record files can be migrated from one datastore to another with Storage vMotion while the VM is running.

Monitoring CPU and memory consumption

VMware provides the ability to monitor resource use in the web interface of VMware vSphere Client. You can monitor CPU usage in the bunch by going to Settings > Monitor > vSphere DRS > CPU Utilization. There are besides other options for monitoring the memory and memory quad for distinguish ESXi hosts. VMware monitoring is supported in NAKIVO Backup & Replication 10.5. Read more about infrastructure monitor in the web log mail.

Using VMware HA and DRS Together

VMware HA and DRS are not competing technologies. They complement each early, and you can use both VMware DRS and HA in a vSphere bunch to provide high handiness for VMs and libra workloads if VMs are restarted by HA on other ESXi hosts. It is recommended that you use both technologies in vSphere clusters running in production environments for automatic pistol failover and load balance.

When an ESXi host fails, VM failover is initiated by HA, and VMs are restarted on other hosts. The first priority in this position is to make VMs available. But after VM migration, some ESXi hosts can be overloaded, which would have a negative affect on VMs running on those hosts. VMware DRS checks the resource usage on each master of ceremonies within a bunch and provides recommendations for the most intellectual placement of VMs after a failover. As a result, you can always be certain that there are enough resources for VMs after failover to run workloads with proper performance. With both VMware DRS and HA enabled you can have a more effective bunch.


VMware provides the herculean functionality of clusters in vSphere to meet the needs of the most demand vSphere customers. This web log stake covered VMware DRS and HA and explained the working principle and chief parameters for each of these clustering features. VMware DRS v HA complement each other and make the final result of using a bunch better. even if you use VMware DRS and HA, wear ’ t forget to back up VMware VMs in vSphere. Download NAKIVO Backup & Replication Free Edition for VMware backing in your environment.

Leave a Reply

Your email address will not be published.