Knowledge
- Explain DRS affinity and anti-affinity rules
- Identify required hardware components to support DPM
- Identify EVC requirements, baselines and components
- Understand the DRS slot-size algorithm and its impact on migration recommendations
Skills and Abilities
- Properly configure BIOS and management settings to support DPM
- Test DPM to verify proper configuration
- Configure appropriate DPM Threshold to meet business requirements
- Configure EVC using appropriate baseline
- Change the EVC mode on an existing DRS cluster
- Create DRS and DPM alarms
- Configure applicable power management settings for ESX Hosts
- Properly size virtual machines and clusters for optimal DRS efficiency
- Properly apply virtual machine automation levels based upon application requirements
Tools
- vSphere Resource Management Guide
- Product Documentation
- vSphere Client
- DRS Resource Distribution Chart
Notes
Explain DRS affinity and anti-affinity rules
- Specifies that two or more virtual machines are placed on the same host or on different hosts.
- When a conflict in rules occurs, the older rule takes precedence and the new rule is disabled.
- Disabled rules are then ignored.
- Higher precedence is given to preventing violations of anti-affinity rules then violations of affinity rules.
To check on affinity rule violations
- Select the cluster in the inventory panel of the vSphere Client
- Select the DRS tab, and click Faults
- Any rule currently being violated has a corresponding fault on this page. Read the fault to determine why DRS is not able to satisfy the particular rule.
Identify required hardware components to support DPM
- Uses IPMI, iLO or WOL
- If one of those three is not supported, DPM can put the host in standby mode.
- If a host supports multiple protocols, the order of precedence is IPMI, iLO, WOL.
- For WOL make sure WOL is supported for physical NICs
Identify EVC requirements, baselines and components
- Configured at Cluster Level
- Helps to ensure vMotion compatibility for the hosts in a cluster by presenting the same CPU feature set to virtual machines.
- Must use same CPU vendors in cluster
- Verify CPU compatibility here VMware KB1003212
- Intel VT or AMD-V
- XD (Execute Disable) or NX (No Execute)
Properly configure BIOS and management settings to support DPM
- See some of the information above. You will want to make sure you properly configure IPMI/iLO as well as WOL if used. WOL will need to be configured and supported for the NIC.
Test DPM to verify proper configuration
- Manually put a host into standby
- Power off VM resources to let DPM take action based on the configured DPM Threshold
Configure appropriate DPM Threshold to meet business requirements
- Similar to DRS, the DPM power on/off recommendations are assigned priorities ranging from 1 to 5 stars.
- Priority 1 is mandatory while priority five will bring just a slight improvement.
- Configured per cluster under Power Management.
- You can override DPM on a host level, specifying particular hosts you would like to only manually involve or disable involvement entirely.
Configure EVC using appropriate baseline
- This VMware KB1003212covers this topic in depth. In it you can find what version of vSphere/ESX support what baselines and what CPUs support what baselines.
Change the EVC mode on an existing DRS cluster
- Cluster Settings->VMware EVC
Create DRS and DPM alarms
Check out this spreadsheet for a complete list of DRS alarms.
http://communities.vmware.com/servlet/JiveServlet/download/12145-1-35516/vSphere%20Alarms%20v2.xlsx
Specifically related to DPM, your most common event to monitor for will be a failure to bring a server back online when it is needed again.
- Exit Standby Error alarm: DrsExitStandbyModeFailedEvent
Additionally these other events exist:
- Entering Standby mode (about to power off host): DrsEnteringStandbyModeEvent
- Successfully entered Standby mode (host power off succeeded): DrsEnteredStandbyModeEvent
- Exiting Standby mode (about to power on the host): DrsExitingStandbyModeEvent
- Successfully exited Standby mode (power on succeeded): DrsExitedStandbyModeEvent
Configure applicable power management settings for ESX Hosts
You can set advanced host attributes to manage power settings.
Power.CpuPolicy: When you set this attribute to the default value of static, VMkernel does not directly set CPU power management states and only
responds to requests from the BIOS. When you enable this policy (set to dynamic), VMkernel dynamically selects appropriate power management states based on current usage. This can save power without degrading performance. Enabling this option on systems that do not support power management results in an error message.
Properly size virtual machines and clusters for optimal DRS efficiency
Realize that not properly sizing your virtual machines will affect the efficiency of DRS. If you give a system too much memory, then less memory is available to other systems. Additionally, the overhead for the system will be higher with a higher amount of configured memory. Ultimately this can result in an increased slot size. One mistake you can make is simply porting systems over with the same specs they had when they were physical, when they don’t need it. If you bring enough multi-core systems over you may see performance issues quickly.
Properly apply virtual machine automation levels based upon application requirements
This will override the cluster settings for that virtual machine.
Under Cluster Settings-> VMware DRS->Virtual machine Options
- Fully Automated(default)
- Partially Automated
- Manual
- Disabled
Other Links