TechOnTip Weblog

Run book for Technocrats

Archive for the ‘MsFailOverClustering’ Category

Application Clustering – Config Build

Posted by Brajesh Panda on May 15, 2011

Before we start building any cluster we may like to use a cheat sheet which will help us to capture all inputs required to build an application cluster.

This list is for reference purpose. You can enhance this list as per your requirement. You can download this sheet from here.

Posted in MsClustering, MsFailOverClustering | Tagged: , | Leave a Comment »

Cluster Validation Error: An error occurred while executing the test. There was an error initializing the network tests. There was an error creating the server side agent (CPrepSrv). Creating an instance of the COM component with CLSID {E1568352-586D-43E

Posted by Brajesh Panda on May 8, 2011

While building failover cluster under one of my friend faced bellow error during validation phase.

An error occurred while executing the test. There was an error initializing the network tests. There was an error creating the server side agent (CPrepSrv). Creating an instance of the COM component with CLSID {E1568352-586D-43E4-933F-8E6DC4DE317A} from the IClassFactory failed due to the following error: 80070005

After some investigation & research on we found this error is applied to Windows 2008 Domain based networks.

Fix:

è Log On to Windows 2008 Domain Controller

è Open DCOMCONFG from start – run

è Expand Component Services -> Computer

è Right Click My Computer -> Select Default Properties Tab

è Under Impersonation Level Default Impersonate & Click Apply

è Reboot Domain Controller & Try Validation Again

Posted in MsClustering, MsFailOverClustering | Tagged: , | 2 Comments »

Cluster Shared Volume ‘volume1’ (‘cluster Disk 1’) is no longer available on this node because of ‘STATUS_BAD_NETWORK_PATH(c0000be)’. All I/O will temporarily be queued until a path to the volume is reestablished.

Posted by Brajesh Panda on April 14, 2011

Symptom: You can access Volume1 from one node but not from other node. This node will through below error in the event log

Cluster Shared Volume ‘volume1’ (‘cluster Disk 1’) is no longer available on this node because of ‘STATUS_BAD_NETWORK_PATH(c0000be)’. All I/O will temporarily be queued until a path to the volume is reestablished.

Solution: Verify if you have forget to enable “Microsoft File & Print Sharing” & “Client for Microsoft Network” features on the CSV network card. This is the requirement for CSV communication.

Mostly we forget to enable these two things when we use traditional cluster private network card for csv communication.

Posted in MsFailOverClustering, MsHyper-V | Tagged: , | Leave a Comment »

SQL Failover Cluster Network Binding Error

Posted by Brajesh Panda on April 21, 2010

1. Remove Ghost Adapters: Mostly happen in virtual machines or where you uninstall & re-install network cards. In this process, you left some Hidden Adapters or GUID in registry. You have to remove them. If it didn’t apply to your environment or didn’t work for you follow second step.

2. Verify Network Binding of “Domain” Adapters using registry. It may be different than the GUI Open in Network Settings Control Panel.

a. Binding are stored inside the registry

Key Name: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesTcpipLinkage

Name: Bind

Type: REG_MULTI_SZ

Data: Device{C7F03A93-3C6E-46F7-B4A8-492C4AF0BD41}

Device{181E62B9-C204-4237-9D9D-6DC16A6CEAE2}

Device{3E9D1781-1418-45A6-B88F-10979F777C15}

Device{3120A369-0C9B-425E-9C84-FE0A54FB3D7B}

Device{78024161-6DA7-4404-9CCA-6687AEC8A73F}

b. Find out to which adapter these GUIDs are referring to “wmic nicconfig get description, SettingID > C:nicconfig.txt

Description SettingID

WAN Miniport (SSTP) {54AB7E97-2074-413D-A578-2119CD23ECB6}

WAN Miniport (L2TP) {C0B9FD38-A0E8-47CA-BF02-5B6D2B663BF7}

WAN Miniport (PPTP) {1D230E0E-4E65-42C2-9859-9B80E1A4E0C5}

WAN Miniport (PPPOE) {E7F554D4-E231-4291-B40E-2031D3F1E55E}

WAN Miniport (IPv6) {A91F2521-CF79-4F98-9E02-6A516B9D496C}

WAN Miniport (Network Monitor) {2131BB57-8C69-4EAE-B5E3-24E4035AB71D}

Broadcom BCM5709C NetXtreme II GigE (NDIS VBD Client) {181E62B9-C204-4237-9D9D-6DC16A6CEAE2}

Broadcom BCM5709C NetXtreme II GigE (NDIS VBD Client) {78024161-6DA7-4404-9CCA-6687AEC8A73F}

WAN Miniport (IP) {7F583D1B-4203-4AD0-A6A0-4BD007D21735}

RAS Async Adapter {7C264673-3322-4312-AE60-D09B4E71E368}

Broadcom BCM5709C NetXtreme II GigE (NDIS VBD Client) #3 {3E9D1781-1418-45A6-B88F-10979F777C15}

Broadcom BCM5709C NetXtreme II GigE (NDIS VBD Client) {3120A369-0C9B-425E-9C84-FE0A54FB3D7B}

Microsoft Failover Cluster Virtual Adapter {C7F03A93-3C6E-46F7-B4A8-492C4AF0BD41}

c. Decision:

i. From Step A & Step B we can see Microsoft Failover Cluster Virtual Adapter with GUID {C7F03A93-3C6E-46F7-B4A8-492C4AF0BD41} is listed on top of the binding order.

ii. As per SQL cluster installation wizard top adapter in the binding sequence must be a DOMAIN adapter & must be able to communicate to production AD Domain.

iii. Generally this value get populated if you have your network card properties is having a DNS suffix value. However in case of Microsoft Failover Cluster Virtual Adapter, this value will not be there because it is a virtual adapter created by failover cluster manager software.

d. To find out a NIC having a DOMAIN attribute value we can verify in the following registry section

i. HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesTcpipParametersInterfaces; It will list down all interface GUIDs & their related attributes & values

ii. You may able to identify your production/public network card (with correct IP address). It’s DOMAIN attribute may have a value or may not be. If there is no value, you can add your DOMAIN SUFFIX in network card properties in GUI or you can edit in registry.

iii. After you get this domain value populated for the public adapter, which is connected to your domain; You will be moving the GUID to the 1st place in the HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesTcpipLinkageBind registry key.

iv. You have to do it for both the nodes.

v. Run Installation After this.

Posted in MsFailOverClustering | Tagged: , | 1 Comment »

SQL 2008 Failover Clustering Guide

Posted by Brajesh Panda on April 21, 2010

Before you start building SQL Cluster; You may like to know what are the things I need to build a cluster. I will advise you read below posts before you start..
  1. Application Clustering – Config Build
  2. Windows Clustering – Reference Architecture & Build
  3. SQL Server prerequisites for installation build a cluster
Let’s Build a SQL 2008 Cluster.  Same steps will also applicable for SQL 2008 R2.

  • SQL 2008 Clustering has been changed a lot from SQL 2000 or SQL 2005 Clustering.
  • Earlier whenever you build SQL clusters all nodes get installed the same time.
  • However incase of SQL 2008 Clustering there are two ways you can complete installation,
    • Standard
      • Create a New SQL Failover cluster node by clicking New SQL Server failover installation
      • Then add other nodes using SQL 2008 Installation Wizard by clicking Add node to SQL Server failover Cluster

  • Advanced
    • Prepare all nodes for failover clustering using Advanced Cluster Preparation one by one. You can use config file (.ini) for quick turnaround.
    • Install complete failover cluster functionality on any node, preferably on the node which is holding the shared disks by clicking Advanced Cluster Completion

Install Microsoft Distributed Transaction Coordinator Cluster Resources

Before installing SQL Server on a failover cluster, determine whether the Microsoft Distributed Transaction Coordinator (MSDTC) cluster resource must be created. If you are installing only the Database Engine, the MSDTC cluster resource is not required. If you are installing the Database Engine and SSIS, Workstation Components, or if you will use distributed transactions, you must install MSDTC. Note that MSDTC is not required for Analysis Services-only instances. You can use a smaller Disk or Quorum Disk for MSDTC itself

  • If you like to create a MSDTC cluster resource, better to make available a Small Disk (Storage) in the system like Quorum. Assign a drive letter to this


  • Select Distributed Transaction Coordinator (DTC) & Click Next


  • Type a name of DTC resource & specify IP Address on Client Access Point window


  • Select Storage Drive for DTC
  • And Finish the wizard

Install 1st Node of SQL 2008 Cluster

  • On SQL Installation wizard, select Installation & then New SQL Failover Cluster Installation
  • It will run through Setup Support Rules testing


  • Verify testing results & press okay to move forward
  • Click Install to Install Support Files
  • It will run Detailed Setup Support Rule Testing for Failover Cluster. Verify the results & move forward
  • Provide License Key & accept End User License Agreement.
  • On features selection page select required features. Note: Reporting Services can’t be clustered. To make reporting service high available we have to use Windows NLB having multiple reporting services.
  • Next Configuration pages depend on the number of features you have selected.

  • On Instance Configuration page type Clustered SQL Server Network Name & Path for SQL Binary installation as Instance root directory. It is the path to local directory. Next window will show you result regarding Disk Space Requirements i.e. if enough space is available for installation.

  • In next Cluster Resource Group window provide SQL Cluster Group name. I used prefer SQL Cluster Network Name – Instance Name

  • Select Storage Drives which will be used for SQL Cluster Resource Group you have created. Not to worry you can change those drives later on. As the 1st drive in the selection will be used for default drive for all databases, I would like to select the bigger drive over here.

  • In Cluster Network Configuration window provide Network IP Address details for SQL Cluster Server

  • On Cluster Security Policy window, you can define security groups. However use service SIDs are recommended

  • On Server Configuration window select Service Accounts. For demonstration purpose I have selected my Domain Admin account over here. But it is absolutely not recommended. I expect you have knowledge on service accounts. This service account don’t need any administrative permission on nodes, minimizing any security risk.

  • On Collation Tab, I would like to retain default config. If you need any specific config please change
  • Database Engine Configuration Page – Account Provisioning page I would like to select Mixed Mode for authentication with a complex password for my SQL SA account.

  • On Data Directories Tab I would like to verify if all paths are pointed my Shared Storage Volume

  • On FILESTREAM tab enable streaming configuration as required

  • On Analysis Services Configuration Window configure account provisioning & data directories. If you have extra shared storage drives you can select those drives over here.
  • On Reporting Services Configuration page select “Install but don’t configure”
  • Configure error reporting as per organization standard. I always ignore them 😉
  • Installation wizard will check Cluster Installation Rules
  • Verify all Summary details & click next for Installation

  • Complete Installation

This installation will install SQL Failover Cluster on the node from where you are running the installation. So after installation it will Create a single node sql failover cluster. So later on we have to join new SQL nodes to make it a high available architecture.

Follow next steps on how to add another node to this Single Node SQL Failover Cluster!

Add Another Node to Existing SQL 2008 Cluster

In my lab I have created a Single Node SQL Cluster & adding additional node later on.

  • Start SQL Installation from Media & select Add node

  • Click Okay
  • Provide Product Key in other window
  • Accept License Agreement
  • Install Support Files
  • Verify Errors & fix it if required.  Even if set the binding order correctly, sometime you may see this error. I have mentioned the steps how to fix it in another post, find it here

  • Next window will show up all cluster config details. Verify everything is okay. Installation wizard detects all config from other nodes which are part of same windows cluster

  • Provide Service Account & Password Details. Installation Wizard picks up all details from other SQL Node in the Fail Over Cluster

  • Configure Error reporting as per your organizations standard
  • Next window will verify all Clustering parameters before it start the Node Joining to the SQL cluster

  • Verify detailed report about the installation & move ahead

Remove a Node from Existing SQL 2008 Cluster

  • You can remove nodes from cluster from Maintenance tab at left hand side of installation wizard & then clicking Remove Node from a SQL Failover Cluster. Start from the Passive Node who don’t won anything in the failover cluster.

After wards it pretty straight forward so you can click next->next-> finish

Posted in MsFailOverClustering | Tagged: , , | 5 Comments »

Best Practices for Exchange 2007 CCR Stretched Cluster

Posted by Brajesh Panda on March 8, 2010

In course of my SI Projects I found these are should be the best practices for a Exchange CCR Stretched Cluster. Let me know if anything I should include in my list.

Decision Making Factors

  • Is automated failover is required in unplanned site outage?
  • How frequently you are facing situation like complete site disaster?
  • Where is the major user base located? Is at Prod Site or DR Site?
  • Do we have enough bandwidth between these two sites?

     

AD Sites

  • Both CCR Hosts must be in same AD Site

    N.B:    Typically, Active Directory communication is controlled based on the AD Site design. Since both datacenters are using the same AD site, there is a risk that servers and clients in 1 datacenter may be communicating to domain controllers in the secondary datacenter. Cross authentication may leads slow response in Directory Server Access!

    –    Plan two domain controllers in each site

Network

  • Encase of Windows 2003 Server we have stretch the subnets across datacenters
  • Windows 2008 Stretch cluster can span across different subnets; however we have to consider below points;
    • Latency must not be more than 50ms
    • Two different subnets should be available for Cluster Private & Public Communication
    • Read the rest of this entry »

Posted in Clustering, Exchange2007, MsClustering, MsFailOverClustering | Tagged: , , , | 1 Comment »

Before you start installing SQL 2005 Cluster

Posted by Brajesh Panda on March 8, 2010

Here is some best practices you should aware of before you start installing Windows Cluster. I will be keep updating this list….. Let me know if I am missing something most important.

Applicable Product: Windows Cluster & SQL 2005

  • Make sure you have installed Advanced/Enterprise/Data Center Edition of Windows OS on all servers
  • Deployment User account is added to local admin groups
  • Deployment User account must have all Active Directory permissions to create Computer accounts for Virtual SQL Node
  • If you’re IT policy is preventing you to get computer account creation rights for deployment user account please go through KB articles how cluster computer accounts can be staged before you start installation
  • Verify Task Scheduler is running on all cluster nodes (For SQL Clustering)
  • Verify Windows Cryptographic Service is running on all cluster nodes (For SQL Clustering)
  • Verify NTLM Security Provider Service is running on all cluster nodes (For SQL Clustering)
  • Verify Remote Registry Service is running without any error on all cluster nodes (For SQL Clustering)
  • Verify you can access media location from other nodes by using UNC path (For SQL Clustering)
  • Apart from SQL Clustering you have to install software individually on all other nodes. Hence make sure all media you are using are of same version & cluster aware.
  1. SQL Clustering use Task Scheduler to install the SQL software on all nodes at the same time.
  2. Apart from SQL Standard Edition, you need Enterprise Edition software for Clustering. SQL Standard edition support only 2 Node clusters
  • Make sure you have the right software media
  • Better to keep handy SMSTrace tool to verify Installation logs encase of any problem
  • File Systems must be formatted as NTFS. Read the rest of this entry »

Posted in Clustering, MsClustering, MsFailOverClustering | Tagged: , | Leave a Comment »

 
%d bloggers like this: