Table of Contents Lab Overview - Components... 2 Business Critical Applications - About this Lab... 3 Infrastructure Components - VMware vcenter... 5 Infrastructure Components - VMware ESXi hosts... 6 Infrastructure Components - Virtual Machines... 7 Application Components - Microsoft SQL server... 8 Infrastructure Components - VMware vsphere App HA... 9 Infrastructure Components - vcenter Operations Manager... 12 Module 1: High Availability for Business Critical Applications... 13 M1.1 - Preparation... 14 M1.2 - Test Policy... 24 M1.3 - Summary... 30 Module 2: CPU Contention Situation on the MS SQL Virtual Machine (CPU Limit in Virtual Machine Settings)... 31 M2.1 - Preparation... 32 M2.2 - Analysis... 38 M2.3 - Remediation... 42 M2.4 - Summary... 47 Module 3: Memory Contention Situation on the MS SQL Virtual Machine (Memory Limit in Virtual Machine Settings)... 48 M3.1 - Preparation... 49 M3.2 - Analysis... 54 M3.3 - Remediation... 58 M3.4 - Summary... 59 Page 1
Lab Overview - Components Page 2
Business Critical Applications - About this Lab VMware vcloud Suite Business Critical Applications Lab This lab demonstrates the use of VMware technology such as vsphere App HA and VMware vcenter Operations Manager to successfully operate and monitor business critical applications (BCAs). The two most important considerations for virtualizing business critical applications are high availability and performance management. This lab provides hands-on experience in using VMware solutions to manage availability and performance for business critical applications. Availability vsphere HA protects against hardware failures, while VMware vsphere vmotion and VMware vsphere Distributed Resource Scheduler (DRS) protect against downtime during planned maintenance activities. These aspects are covered in lab HOL-SDC-1310. These features protect business critical applications during unplanned and planned hardware related downtimes. Many business critical applications have single points of failure such as load balancers, central queue and lock management services, and so on, that have a small compute footprint and can fit within one vcpu. These critical components can leverage VMware vsphere Fault Tolerance for protection and instantaneous always-on recovery. To protect against application level failures, there is a need for an app aware monitoring and remediation platform. vsphere APP HA with vfabric Hyperic provides application aware monitoring and remediation. This lab will provide hands-on experience with creation of application protection policies and remediation. Performance Management Performance management is critical for business critical applications. The ability to monitor, troubleshoot and resolve performance related problems quickly is very important. vcenter operations provides the capability to monitor, troubleshoot, and manage performance problems. In this lab vcenter Operation Managers will be showcased to troubleshoot and perform a quick root cause analysis on a CPU related problem and a memory related problem. The traditional IT manager dashboard will be used for infrastructure visibility. The SQL dashboard that comes with the vcenter Operations Management pack will be used to look at database level performance criteria. Modules in this lab Page 3
Module 1: High Availability for Business Critical Applications Module 2: CPU Contention Situation on the MS SQL Virtual Machine (CPU limit in virtual machine settings). Module 3: Memory Contention Situation on the MS SQL virtual machine (memory limit in virtual machine settings). Page 4
Infrastructure Components - VMware vcenter In this lab VMware vcenter manages a 3-node VMware vsphere Cluster and several virtual machines. Overview of the vcenter Environment Used for the VMware vcloud Suite BCA Lab This screenshot shows that a relatively small environment is used for this lab. VMware vcenter Server is built using the appliance. Page 5
Infrastructure Components - VMware ESXi hosts Overview of VMware ESXi resources used. vsphere Hosts The lab consists of three ESXi servers. Page 6
Infrastructure Components - Virtual Machines Listing of All Virtual Machines Contained in the Lab Setup The lab is built on the following four virtual machines: One Microsoft Windows 2008 R2 running SQL Server 2008R2. Two SLES 11 virtual machines running tc Server 2.8. One SLES 11 virtual machines running htproxy as an HTTP load balancer. Page 7
Application Components - Microsoft SQL server Microsoft SQL Server Management Studio This lab uses Microsoft SQL Server 2008 R2 to host the databases that run the Spring Trader application. Page 8
Infrastructure Components - VMware vsphere App HA Page 9
vsphere App HA Helps Minimize Downtime from Application Failure vsphere App HA is a new feature that complements the functionality of VMware vsphere High Availability with application-level monitoring and automated remediation. App HA improves application uptime through: Policy-based application monitoring and remediation. Integration with vsphere HA to provide multiple recovery options (such as restarting the application service or restarting the virtual machine). Reduces Complexity App HA eliminates the need for application-specific clustering solutions, reducing the complexity involved in managing infrastructure silos and the associated cost. App HA provides: Uniform protection and support for many commonly used, off-the-shelf applications. Simple, centralized management through app-health visibility from vcenter Server. Architecture Overview vsphere App HA leverages VMware vcenter Hyperic to monitor applications. Deploying App HA begins with provisioning two virtual appliances (per vcenter Server): App HA and vcenter Hyperic. The App HA virtual appliance stores and manages App HA policies, and vcenter Hyperic monitors applications and enforces App HA policies. After the simple process of deploying the vcenter Hyperic and App HA virtual appliances, vcenter Hyperic agents are installed in the virtual machines containing applications that will be protected by App HA. App HA Policies App HA policies are easily configured in the Administration section of the vsphere Web Client. Policies define items such as the number of times App HA will attempt to restart a service, the number of minutes it will wait for the service to start, the option to reset the virtual machine if the service fails to start, and the option to reset the virtual machine when the service is unstable. Page 10
Page 11
Infrastructure Components - vcenter Operations Manager Introduction to vcenter Operations Manager vcenter Operations Manager is the key component of the vcenter Operations Management Suite. It provides a new and simplified approach to operations management of vsphere physical and cloud infrastructure. Using patented, self-learning analytics and an open, extensible platform, vcenter Operations Manager includes operations dashboards that provide insights and visibility into health, risk and efficiency of infrastructure, performance management, and capacity optimization capabilities. Using vcenter Operations Manager, you can do the following: Gain comprehensive visibility into the health, risk, and efficiency of your infrastructure and applications. Proactively manage the health of vsphere, virtual machines, and applications. Spot potential performance bottlenecks early and remediate before they affect end users. Right-size and reclaim overprovisioned capacity to increase consolidation ratios. Manage thousands of virtual machines, physical servers, and applications across multiple data centers from a single console. Automatically correlate and analyze monitoring data across infrastructure and applications silos to gain a holistic view of root cause and effect. vcenter Operations Manager and the vcenter Operations Management Suite are essential tools for infrastructure teams and operations teams who want to enable quality of service, operational efficiency, and continuous compliance of dynamic cloud infrastructure and business critical applications. Page 12
Module 1: High Availability for Business Critical Applications Page 13
M1.1 - Preparation ACME requires that the SQL database server for its business critical application be highly available. There is a requirement for SQL to be protected against hardware and application failures. Log In to the vcenter Web Client Log in to the vsphere Web client: User name: administrator Password: VMware1! Page 14
Go to the Administration Menu and Select vsphere App HA Page 15
Create a New Policy by Clicking + Page 16
Select SQL 2008 R2 as the Application Service Page 17
Set Up Remediation Parameters for the Service As Shown Page 18
Ignore Alarm Setup and Click Next Page 19
Finish Creating Policy Page 20
The New Policy is Displayed Page 21
Enable Policy The created policy needs to be assigned to services for it to be enabled. Browse to the vcenter server and click the Monitor tab. Select vcenter App HA. All the compatible applications for App HA in the environment are displayed. Select the MS SQL 2008 R2 and then right-click and select Assign Policy. Page 22
Assign Policy to MS SQL Application Page 23
M1.2 - Test Policy This section shows how to verify that newly created App HA policy is working. Use Remote Desktop to Access SQL Server Page 24
Stop the SQL Service Launch the services view by clicking Start and entering services.msc in the prompt. Browse to the SQL server service. Right-click and select Stop. Page 25
View the App HA Alert in vcenter Hyperic Console Log in to the vcenter Hyperic console and look at the dashboard. The dashboard will show an alert, and the SQL server is automatically restarted. Application HA Maintenance Mode (Optional) Browse to Application Availability under the Monitor tab in vcenter and put the MS SQL 2008 R2 service in maintenance mode. Manually Stop the SQL Service You can now go back to your RDP session for SQL Server and shut down the SQL service again. Page 26
Wait for a few minutes and confirm that the SQL service is not restarted. This shows that if a service is in App HA maintenance mode it can be brought down without App HA reacting to it. Page 27
Restart SQL Server Restart SQL Server manually from the services console. Page 28
Exit App HA Maintenance mode Browse to Application Availability under the Monitor tab vcenter and exit maintenance mode for the MS SQL R2 server instance. Right-click and select Exit Maintenance Mode. Page 29
M1.3 - Summary This module demonstrated the workings of vsphere App HA and its ability to protect from application level failures. vsphere App HA works with vcenter Hyperic to monitor and remediate supported applications. vsphere App HA is simple to configure and is set up with a few mouse clicks. Using App HA along with vsphere HA and vsphere Fault Tolerance can protect business critical applications from hardware and software related failures. vsphere vmotion and DRS help address planned downtime. The VMware HA components improve SLA and availability metrics for business critical applications, with reduced planned and unplanned downtime. Page 30
Module 2: CPU Contention Situation on the MS SQL Virtual Machine (CPU Limit in Virtual Machine Settings) Page 31
M2.1 - Preparation This use case puts parts of the application into a resource contention situation for CPU resources. This simulates a scenario in which the Microsoft SQL Server does not get sufficient CPU resources and shows how this issue appears when vcenter Operations Manager is used. Page 32
Generate Resource Contention Situation - Log In to vsphere Web Client Note: For your convenience, you can use the file named Cheat Sheet.txt to cut and paste commands and other information used in this lab. This file is located on the desktop of the ControlCenter. Using the bookmark shortcuts in Internet Explorer, add a new tab and choose the bookmark for vsphere Web Client. Log in as follows: User name: administrator Password: VMware1! Page 33
Browse to the MS SQL Server Virtual Machine Click VMs and Templates. Page 34
Browse to the MS SQL Server Virtual Machine (cont.) 1. Click Related Objects. 2. Click Virtual Machines. 3. Click sql-w8-01a. Page 35
Manage the MS SQL Server Virtual Machine 1. Click Manage. 2. Click Edit. Page 36
Set CPU Limit 1. Click to expand the virtual machine CPU settings. 2. Change Limit from Unlimited to 250 MHz. 3. Click OK. Note: It takes a while for the effects to be visible in vcenter Operations Manager or vcenter Hyperic. Page 37
M2.2 - Analysis ACME is running into a resource contention situation on Microsoft SQL Server. Observe the IT Manager dashboard to see whether this can be observed. Note: It can take a few minutes for the actual behavior to be shown in the UI. Log In to vcenter Operations Using the bookmark shortcuts in Internet Explorer add a new tab and choose the bookmark for Custom UI vcenter Operations. Log in with credentials: User name: admin Password: VMware1! Page 38
Observe CPU Alerts in vcenter Operations Select the IT Manager tab on the vcenter Operations Manager screen. Sort the active alerts in descending order by start time. Locate the SQL Server alert and double-click. Examine the Alerts More Closely Under the Root Cause tab, explore the message event and the related change event that caused this alert. This is useful to correlate alerts to potential change events in the environment. Page 39
Troubleshooting Click Troubleshoot(see previous screenshot) and select Mashup to look at graphical details of the alert for analysis. Page 40
Alert Metrics and Anomalies Go to the Metric Charts tab which shows the anomaly relating to the alert with CPU Ready above normal threshold. Page 41
M2.3 - Remediation Remediate Resource Contention Situation - Log In to vsphere Web Client Log in as follows: User name: Administrator Password: VMware1! Page 42
Browse to the MS SQL Server Virtual Machine Click VMs and Templates. Page 43
Browse to the MS SQL Server Virtual Machine (cont.) 1. Click Related Objects. 2. Click Virtual Machines. 3. Click sql-w8-01a. Page 44
Manage the MS SQL Server Virtual Machine 1. Click Manage. 2. Click Edit. Page 45
Remove CPU Limit 1. Click to expand the virtual machine CPU settings. 2. Change Limit from 250 MHz to Unlimited. 3. Click OK. Note: It takes a while for the effects to be visible in vcenter Operations Manager or vcenter Hyperic. Page 46
M2.4 - Summary This module demonstrated how to use vcenter operations for troubleshooting CPU related problems at the VM level and performing root cause analysis that ties it back to changes made in the environment. vcenter Operations Manager also provides metrics and anomaly data that speed up the troubleshooting process. This is important for business critical applications to avoid or minimize downtime. This also demonstrates that setting limits on CPU can be dangerous. Page 47
Module 3: Memory Contention Situation on the MS SQL Virtual Machine (Memory Limit in Virtual Machine Settings) Page 48
M3.1 - Preparation Generate Resource Contention Situation - Log In to vsphere Web Client Note: For your convenience, you can use the file named Cheat Sheet.txt to cut and paste commands and other information used in this lab. This file is located on the desktop of the ControlCenter. Log in as follows: User name: administrator Password: VMware1! Page 49
Browse to the MS SQL Server Virtual Machine Click VMs and Templates. Page 50
Browse to the MS SQL Server Virtual Machine (cont.) 1. Click Related Objects. 2. Click Virtual Machines. 3. Click sql-w8-01a. Page 51
Manage the MS SQL Server Virtual Machine 1. Click Manage. 2. Click Edit. Page 52
Configure Memory Limit to Simulate a Resource Contention Issue 1. Click Memory to expand. 2. For Limit,enter 250 MB and enter 0 for Reservation. 3. Click OK. Page 53
M3.2 - Analysis ACME is experiencing memory resource contention on Microsoft SQL Server. Note: It can take a few minutes for the actual behavior to be shown in the UI. Observe the IT Managers Dashboard Browse to the Alerts Overview tab to look at the initial alerts generated. Alerts for MemoryContention are displayed in the Active Alerts pane.double-click the one of the memory related alerts for the SQL server. Page 54
Detailed View of the Memory Alert The more detailed view shows that multiple KPIs are breaching for memory contention in the mashup pane. Page 55
Root Cause Analysis Further examining the change events shown in the Root Cause pane, you will eventually find the resource allocation change you did in the preparation section. Page 56
What Can the SQL Administrator Observe? Select the SQL Server Overview tab on the vcenter Operations Manager screen. This dashboard shows how the SQL administrator can observe the issue surfacing. Due to the memory contention, swapping would go up (2). This leads to higher disk I/O (1) and higher CPU usage (3). Page 57
M3.3 - Remediation The resource contention can be remediated by reversing the change regarding memory limits. Note: Reversing the change in the context of this lab will not necessarily produce any results in a timely manner. This is because it could take hours for the pages that have been swapped out due to the memory shortage to go back into real memory based on the usage characteristics of the SQL server. Page 58
M3.4 - Summary This module showed how to use vcenter operations for troubleshooting memory related problems at the VM level and at the SQL server application level. The SQL Server Management pack for vcenter Operations Manager provides the ability to collect and look at SQL level metrics from vcenter Operations Manager and correlate events across the entire infrastructure stack. Application specific management packs for vcenter Operations Manager can help the infrastructure and application team members collaborate in problem resolution. This module also demonstrated that setting limits on memory can be dangerous. Page 59
Conclusion Thank you for participating in the VMware Hands-on Labs. Be sure to visit http://hol.vmware.com/ to continue your lab experience online. Lab SKU: Version: 20141126-112055 Page 60