Microsoft KB Archive/243195

= Event ID 1034 for MSCS Shared Disk After Disk Replacement =

Article ID: 243195

Article Last Modified on 11/1/2006

-

APPLIES TO


 * Microsoft Windows NT Server 4.0 Enterprise Edition

-



This article was previously published under Q243195



IMPORTANT: This article contains information about modifying the registry. Before you modify the registry, make sure to back it up and make sure that you understand how to restore the registry if a problem occurs. For information about how to back up, restore, and edit the registry, click the following article number to view the article in the Microsoft Knowledge Base:

256986 Description of the Microsoft Windows Registry



SYMPTOMS
After you replace a failed hard disk, or change drives (different SCSI ID or physical location in a SCSI rack) for the shared disk resource, the Cluster service may not start. Also, the following error message may be displayed in the Event log:

Event ID: 1034

Source: ClusDisk

Description: The disk associated with cluster disk resource  could not be found. The expected signature of the disk was.



CAUSE
This issue can occur because the Microsoft Cluster Server (MSCS) server relies on disk signatures to identify and mount volumes. If a hard disk is replaced or the bus is re-enumerated, MSCS may not find the disk signatures that it is expecting, and consequently may fail to mount the disk.



RESOLUTION
WARNING: If you use Registry Editor incorrectly, you may cause serious problems that may require you to reinstall your operating system. Microsoft cannot guarantee that you can solve problems that result from using Registry Editor incorrectly. Use Registry Editor at your own risk.

To resolve this issue:  Back up the Servers Configuration:  Update your Emergency Repair disk (ERD) on each node. To do so, click Start, click Run, type rdisk, and then click OK.  Disable the Cluster Service on both Nodes:  On both nodes, change the Startup value to Manual for the following items:  Clusdisk (by using the Devices tool in Control Panel)</li> Cluster Service (by using the Services tool in Control Panel)</li></ul>

Note Do not try to stop the Clusdisk resource. If you do, you may receive the following error message:

Error 2191: The requested pause or stop is not valid for this service

</li> Restart node 2.  At the OS Loader Boot menu, press SPACEBAR to pause the boot process and prevent Windows NT from loading.</li></ul>

Note This is necessary to maintain termination on the shared SCSI bus on some hardware configurations.</li> Restart node 1 (and allow it to boot completely).

Note It is assumed at this point that you have replaced the failed hard disk in the shared resource. It is also assumed that you have followed the necessary steps to configure the hard disk to permit Windows NT to recognize it upon restarting.</li></ol> </li> Configure the New Disk: <ol style="list-style-type: lower-alpha;"> Start Disk Administrator (Windisk.exe) on node 1.</li> Partition the new drive.</li> Format the partition with the NTFS file system.</li> Assign a drive letter and label to the partition.</li></ol> </li> Gather Information about your Disks: <ol style="list-style-type: lower-alpha;"> Start FTEdit.exe (included in the Microsoft Windows NT or 2000 Resource Kit).</li> Obtain the disk signature information for all the hard disks.</li> Document which disk signature belongs to each partition label.

Note At this point, there are probably 2 disk signatures for the hard disk that you are replacing. For example, if you are replacing disk W:, you see two separate entries in Ftedit for disk W:, each with its own disk signature. One signature is the original disk, the second signature is the new replacement.</li></ol>

For additional information about the Ftedit.exe tool, click the article number below to view the article in the Microsoft Knowledge Base:

149927 FTedit.exe: What You Can and Can't Use It For

</li> Gather Information about your Disk Resources: <ol style="list-style-type: lower-alpha;"> <li>Start the Registry Editor (Regedt32.exe).</li> <li>Open the HKEY_LOCAL_MACHINE hive, and then click on the root (HKEY_LOCAL_MACHINE) to select it.</li> <li>On the Registry menu, click Load Hive.</li> <li>Locate the %SystemRoot%\Cluster folder, and then click CLUSDB. When you are prompted for a key name, type Cluster .</li> <li>Locate the following registry key:

<ul> <li>Under each GUID, look for a TYPE value that is Physical Disk. After you identify all the physical disks, verify the  key. It should have a value of Disk . This is how you can identify which GUID belongs to which physical disk resource.</li></ul>

Note: Each resource is listed by its Globally Unique Identifier (GUID). The GUID has 32 alpha numeric values with the following format: x11xx1x1-x11x-11x1-xxx1-11111xxx111x.</li> <li>Document which disk signature belongs to each physical disk resource.</li></ol> </li> <li>Update the Disk information in the Cluster Registry: <ol style="list-style-type: lower-alpha;"> <li>Compare the disk signatures for each hard disk that you observed using the FTEdit.exe tool (step 4) with the disk resource signature values in the Cluster registry (step 5).</li> <li>The disk signature from FTEdit that is not the same as the one listed in the Signature key is the correct new disk signature.</li> <li>Document the correct disk signature for each disk.</li> <li>Replace the information in the Signature key in the Cluster registry with the correct disk signature value for the new disk. Make sure to make the changes in Hex.</li> <li>Locate the following registry key for the disk that is being replaced:

Note The data type for the DiskInfo key should be REG_BINARY, and should be located directly above the  key that was modified earlier in this article.</li> <li>Delete the  registry key.

Note Make sure to delete the entire key, and not just the value for the key. This key is dynamically re-created the next time the Cluster service is started.</li> <li>Select the Cluster hive, then on the Registry menu, click Unload Hive.</li></ol> </li> <li>Remove the Disk Signature Information for the Cluster Disk driver: <ol style="list-style-type: lower-alpha;"> <li>Locate and delete the following registry key:

Note There are several registry keys under the  key. They are disk signatures for drives on the shared resource. This key and all subkeys are re-created the next time the Cluster service bring that disk resource online on that node.</li> <li>Quit Regedt32.exe.</li></ol> </li> <li>Remove the Duplicate entry in the Disk Key: <ol style="list-style-type: lower-alpha;"> <li>Start FTEdit.exe again.</li> <li>Locate and then click the disk signature for the old disk (determined in step 6).</li> <li>Click Edit, and then click Delete Drive.</li> <li>Quit Ftedit.exe, and then click Yes to save changes.</li></ol> </li> <li>Create a Copy of the Cluster Registry: <ul> <li>Copy the Clusdb file located in the %SystemRoot%\Cluster folder to a floppy disk.</li></ul> </li> <li>Remove the Cluster Registry on the Quorum drive: <ul> <li>Start Windows Explorer, go to the MSCS folder on the quorum drive and delete the Quolog.log file and all the Chk .tmp files.

Note The Cluster registry is dynamically re-created on the Quorum drive when the Cluster Service starts.</li></ul> </li> <li>Start the Cluster Service on Node 1: <ol style="list-style-type: lower-alpha;"> <li>Change the following Startup values back to: <ul> <li>System for Clusdisk (by using the Devices tool in Control Panel)</li> <li>Automatic for the Cluster Service (by using the Services tool in Control Panel)</li></ul>

</li> <li>Manually start the Clusdisk driver.</li> <li>Manually start the Cluster service.Important If the quorum drive is the drive that failed, you may need to start the Cluster service by using the /fixquorum switch and temporarily change the drive that is the quorum drive.</li> <li>Verify on node 1 that Clusdisk and the Cluster service started correctly. Start Cluster Administrator and verify that all physical disk resources are online and that there is one for each volume on the shared disk.</li></ol> </li> <li>Boot Node 2: <ul> <li>Keep node 1 on, and then start node 2.

Reminder Node 2 will fail to join the cluster because the Cluster Service is disabled.</li></ul> </li> <li>Manually Synchronize the Cluster Registries: <ul> <li>Replace the %SystemRoot%\Cluster\Clusdb file on node 2 with the Clusdb file that you copied to the floppy from node 1.

Important: After the cluster service starts on node 1, it is impossible to copy this file without stopping the Cluster service and unloading it from the registry.</li></ul> </li> <li>Remove the Disk Signature Information for the Cluster Disk driver on node 2: <ul> <li>Locate and delete the following registry key:

</li></ul> </li> <li>Remove the Duplicate entry in the Disk Key: <ol style="list-style-type: lower-alpha;"> <li>Start FTEdit.exe again.</li> <li>Locate and then click the disk signature for the old disk (determined in step 6).</li> <li>Click Edit, and then click Delete Drive.</li> <li>Quit Ftedit.exe, and then click Yes to save changes.</li></ol> </li> <li>Start the Cluster Service on Node 2: <ol style="list-style-type: lower-alpha;"> <li>Change the following Startup values back to: <ul> <li>System for Clusdisk (by using the Devices tool in Control Panel)</li> <li>Automatic for the Cluster service (by using the Services tool in Control Panel)</li></ul>

</li> <li>Manually start the Clusdisk driver.</li> <li>Manually start the Cluster service.</li> <li>Verify that node 2 successfully joined the cluster by using Cluster Administrator. Verify that you are able to manually move the new hard disk resource back and forth between the two nodes. To do so, right-click the resource group the disk is in and then click Move Group. This option changes ownership of the resource group and all its contents to the other node.</li></ol> </li></ol>

Perform the Following Steps if the Previous Steps Did Not Resolve the Issue
In the event that the steps listed earlier in this article do not resolve your issue, you may have corruption in the Disk key and it may be necessary to delete it from the registry. The Disk key is dynamically re-created when Disk Administrator is launched for the first time. The following procedure outlines the steps you need to perform in addition to the previous steps.

Note This process can take a significant amount of time and should only be used if the previous procedure does not resolve your issue.

Warning Verify that there is no software fault tolerance on the systems. Deleting the Disk key from the registry destroys all software fault-tolerant sets, and all data on them is lost! Software fault tolerance includes disk mirroring, disk duplexing, stripe sets with parity, and stripe sets without parity. Any software fault tolerance on the system is local to the nodes in addition software fault tolerance is not supported on cluster shared drives. <ol> <li>Back up your Configuration on Both Nodes: <ol style="list-style-type: lower-alpha;"> <li>Start Disk Administrator (Windisk.exe), and save a copy of the disk configuration information for disaster recovery. To do so, click Configuration Save in the Partition drop-down list, and then quit Disk Administrator.</li> <li>Update your Emergency Repair disk (ERD). To do so, click Start, click Run, type rdisk, and then click OK.</li></ol> </li> <li>Disable the Cluster Service on both Nodes: <ol style="list-style-type: lower-alpha;"> <li> On both nodes change the <pre class="fixed_text">Startup value to Manual for the following items: <ul> <li>Clusdisk (by using the Devices tool in Control Panel)</li> <li>Cluster Service (by using the Services tool in Control Panel)</li></ul>

Note Do not try to stop the Clusdisk resource. If you do, you may receive the following error message:

Error 2191: The requested pause or stop is not valid for this service

</li> <li>Restart node 2:

At the OS Loader Boot menu, press the SPACEBAR to pause the boot process and prevent Windows NT from loading.

Note This is necessary to maintain termination on the shared SCSI bus on some hardware configurations).</li> <li>Restart node 1 (and allow it to boot completely)

.Note It is assumed at this point that you have replaced the failed hard disk in the shared resource. It is also assumed that you have followed the necessary steps to configure the hard disk to permit Windows NT to recognize it upon restarting.</li></ol> </li> <li>Label All Partitions: <ol style="list-style-type: lower-alpha;"> <li>Start Windows Explorer.</li> <li>Right-click the drive, and then click Properties.</li> <li>Type the partition drive letter in the Disk Label box under the General tab, and then click OK.

Note Label all hard disks with their current drive letters. For example, if disk R: is labeled "SQL Data", change the label so that disk R: is labeled "Disk R". This prevents confusion later after deleting the Disk key.</li></ol> </li> <li>Re-create your Disk Configuration: <ol style="list-style-type: lower-alpha;"> <li>Start the Registry Editor (Regedt32.exe).</li> <li>Locate and then delete the following registry key:

</li> <li>Reboot node 1.</li> <li>Start Disk Administrator. You may receive a warning message that Disk Administrator is being run for the first time on this computer, and this message is expected.</li> <li>Change all partitions so that their drive letters match their labels.

Note It may be necessary to restart the computer several times in order to get all the drive letters changed back to the way they were.</li> <li>Partition and format the new disk and assign the same drive letter to it that the previous (failed) hard disk had.</li> <li>After all drive letters are correct, quit Disk Administrator.</li></ol> </li> <li>Gather Information about your disks: <ol style="list-style-type: lower-alpha;"> <li>Start Ftedit.exe.</li> <li>Document all of the disk signatures and their corresponding drive letters.</li> <li>Quit FTEdit.exe.</li></ol> </li> <li>Update the Disk information in the Cluster Registry: <ol style="list-style-type: lower-alpha;"> <li>Start the Registry Editor (Regedt32.exe).</li> <li>Open the HKEY_LOCAL_MACHINE hive, and then click on the root (HKEY_LOCAL_MACHINE) to select it.</li> <li>On the Registry menu, click Load Hive.</li> <li>Locate the %SystemRoot%\Cluster folder, and then click CLUSDB. When you are prompted for a key name, type Cluster .</li> <li>Locate the following registry key:

<ul> <li>Under each GUID, look for a  key that has a Physical Disk value. After you identify all the physical disks, verify the  key. It should have a value of Disk . This is how you can identify which GUID belongs to which physical disk resource.</li></ul>

Note Each resource is listed by its Globally Unique Identifier (GUID). The GUID has 32 alpha numeric values and will follow this format: x11xx1x1-x11x-11x1-xxx1-11111xxx111x.</li> <li>Compare the disk signatures for each disk that you observed by using the Ftedit.exe tool with the Signature values in the Cluster registry. When you locate the hard disk that you are replacing, make a note of the old signature value (you will need it later). Replace the disk signature information with the new hard disk's signature that you obtained from Ftedit.exe.</li> <li>Locate the following registry key for the hard disk that you replaced:

<ul> <li>The data type for the DiskInfo key is REG_BINARY. This key should be directly above the  key that you modified earlier in this article.</li></ul> </li> <li>Delete the DiskInfo key. This key is dynamically re-created the next time the cluster service is started.</li> <li>Select the Cluster hive, then on the Registry menu, click Unload Hive.</li></ol> </li> <li>Remove the Disk Signature Information for the Cluster Disk driver: <ol style="list-style-type: lower-alpha;"> <li>Locate and delete the following registry key:

NOTE: There are several registry keys under the Signatures key. They are disk signatures for drives on the shared resource. This key and all subkeys are re-created the next time the Cluster service bring that disk resource online on that node.</li> <li>Quit Regedt32.exe.</li></ol> </li> <li>Remove the Cluster Registry on the Quorum drive: <ul> <li>Start Windows Explorer, go to the MSCS folder on the quorum drive and delete the Quolog.log file and all the Chk .tmp files.Note: The Cluster registry will be dynamically re-created on the Quorum drive when the Cluster Service starts. For additional information, click the article number below to view the article in the Microsoft Knowledge Base:

280353 How to Change Quorum Disk Designation

</li></ul> </li> <li>Create a Copy of the Cluster Registry: <ul> <li>Copy the Clusdb file located in the %SystemRoot%\Cluster folder to a floppy disk.</li></ul> </li> <li>Start the Cluster Service on Node 1: <ol style="list-style-type: lower-alpha;"> <li>Change the following Startup values back to: <ul> <li>System for Clusdisk (by using the Devices tool in Control Panel).</li> <li>Automatic for the Cluster Service (by using the Services tool in Control Panel).</li></ul>

</li> <li>Manually start the Clusdisk driver.</li> <li>Manually start the Cluster service.Important If the quorum drive is the drive that failed, you may need to start the Cluster service by using the /fixquorum switch and temporarily change the drive that is the quorum drive. For additional information, click the article number below to view the article in the Microsoft Knowledge Base:

280353 How to Change Quorum Disk Designation

</li> <li>Verify on node 1 that Clusdisk and the Cluster service started correctly. Start Cluster Administrator and verify that all physical hard disk resources are online and that there is one for each volume on the shared disk.</li></ol> </li> <li>Boot Node 2: <ol style="list-style-type: lower-alpha;"> <li>Restart node 1. At the OS Loader Boot menu, press SPACEBAR.</li> <li>On node 2, select the OS from Boot Menu and allow it to boot.</li></ol> </li> <li>Re-create your Disk Configuration: <ol style="list-style-type: lower-alpha;"> <li>Start the Registry Editor (Regedt32.exe).</li> <li>locate and then delete the following registry key:

</li> <li>Reboot node 2.</li> <li>Start Disk Administrator. You may receive a warning message that Disk Administrator is being run for the first time on this computer, and this message is expected.</li> <li>Change all partitions so that their drive letters match their labels.

Note It may be necessary to restart the computer several times in order to get all the drive letters changed back to the way they were.</li> <li>Partition and format the new disk and assign the same drive letter to it that the previous (failed) hard disk had.</li> <li>After all drive letters are correct, quit Disk Administrator.</li></ol> </li> <li>Manually Synchronize the Cluster Registries: <ul> <li>Replace the %SystemRoot%\Cluster\Clusdb file on node 2 with the Clusdb file that you copied to the floppy from node 1.Important After the cluster service starts on node 1, it is impossible to copy this file without stopping the Cluster service and unloading it from the registry.</li></ul> </li> <li>Remove the Disk Signature Information for the Cluster Disk driver: <ul> <li>Start Regedt32.exe, locate and then delete the following registry key:

</li></ul> </li> <li>Start the Cluster Service on Node 2: <ol style="list-style-type: lower-alpha;"> <li>Change the following Startup values back to: <ul> <li>System for Clusdisk (in the Devices tool in Control Panel)</li> <li>Automatic for the Cluster service (in the Services tool in Control Panel)</li></ul>

</li> <li>Manually start the Clusdisk driver.</li> <li>Manually start the Cluster service.Important: If the quorum drive is the drive that failed, you may need to start the Cluster service by using the /fixquorum switch and temporarily change the drive that is the quorum drive. For additional information, click the article number below to view the article in the Microsoft Knowledge Base:

243195 How to Change Quorum Disk Designation

</li> <li>Verify on node 2 that Clusdisk and the Cluster service started correctly. Start Cluster Administrator and verify that all physical hard disk resources are online and that there is one for each volume on the shared disk.</li></ol> </li> <li>Start Node 1 and allow it to join the Cluster: <ol style="list-style-type: lower-alpha;"> <li>Select the OS from Boot Menu and allow node 1 to boot.</li> <li>Verify that node 1 successfully joined the cluster by using Cluster Administrator. Verify that you can manually move the replaced hard disk resource back and forth between the nodes. To do so, right-click a resource group, and then select Move Group.</li></ol> </li></ol>

<div class="references_section">