Necron

Patch Management der Hosts würde ich ab Windows Server 2012 nicht mehr über den SCVMM machen. Viel zu aufwendig mit den Baselines, da ist das CAU deutlich komfortabler. Bare-Metal Deployment ist ein Punkt wo es nicht so abwägig ist. Aber wie immer Geschmackssache. Für mich würde bei einem separaten Managementcluster, worauf nur System Center läuft, auch ein separates Storage im Hintergrund dazu zählen. Ansonsten kann ich die Managementtools auch mit in den Produktiv Cluster nehmen.

Hi, selbst eine Live Storage Migration der VMM VM, die vom VMM initiiert wurde funktioniert. Ebenso eine Shared Nothing Live Migration.

Hi, wenn man den NLB einsetzt und dazu Hyper-V sollte man in der Netzwerkkonfiguration der VM das MAC Spoofing aktivieren.

Das stimmt, aber Starwind kann zwischen zwei iSCSI Targets die LUNs synchronisieren, dass kann das Windows iSCSI Target nicht.

Hi und willkommen an Board, die Konfiguration passt. Auch die Netzwerkkonfiguration ist soweit stimmig. Hier könnte man allerdings eine der 3x NICs aus dem Team herausnehmen und zusätzlich für den Live Migration Traffic verwenden. Switch Independent Teaming mit dem Load Balancing Mode Dynamic ist die Best Practice Empfehlung von Microsoft, man kann es Richtung LACP ändern muss es aber nicht. Das einzige was ich etwas merkwürdig finde, wieso die Windows 7 VMs als Datenbankserver fungieren?

Kurze Info für euch: Die Vorträge auf der cim werden nächste Woche veröffentlicht.

Hi, Du willst also Virtual to Physical machen? Wie zahni schon sagte vergiss es. P2V geht, aber V2P würde ich im Leben nicht machen.

Hallo Community, seit heute steht das Update 3 für das Windows Azure Pack zur Verfügung. Damit werden die folgenden 10 Probleme behoben: For stand-alone virtual machines (VMs) whose templates specified multiple network adapters, the tenant user cannot associate a particular network to each network adapter. Symptom: If the user is creating a VM that’s based on a template that contains multiple network adapters, the Tenant Site VM extension will automatically assign the first available network. Resolution: The tenant user now has the necessary user interface (UI) to select a particular network for a particular network adapter. Developers who use Web Site Cloud to deploy PHP applications do not have an option to specify PHP 5.5 version. Symptom: If the user wants to deploy a PHP 5.5 application, there is no option to specify that version of PHP. Resolution: The user now has the PHP 5.5 version option. Usage records that contain more than 4,000 characters of resource data are not supported. Symptom: If the Usage Collector encounters a usage record that has more than 4,000 characters, it fails and shuts down the usage pipeline for System Center usage data. Resolution: Some of the instructions in this article involve running an SQL script that will increase the size of the table field to nvarchar(max). Tenants experience unexpected behavior in the portal when they create websites that have the same name in different plan subscriptions that are each on a separate Web Site cloud. Symptom: When a tenant creates two websites that have the same name on different Web Site clouds and subscriptions, the portal may display inaccurate information about one or both of the websites. Resolution: Users can now create websites that have the same name through separate subscriptions and clouds without seeing incorrect information about their sites or subscriptions. Tenants who create an ASP.NET site by using Web Site Cloud cannot specify whether the Application Pool will be running in Classic or Integrated Mode. Symptom: When a tenant tries to create a website, the user notices that the Application Mode setting is not available. If the application requires Classic Mode, it will run the risk of failing at run time. Resolution: Users can now specify the Application Pool mode for their applications. Administrators cannot use Fully Qualified Domain Names (FQDN) when they configure SQL Always-On Listeners, because the Windows Azure Pack Resource Provider for SQL Server does not support it. Symptom: When an administrator enters an FQDN in the SQL Extension of the Administration site, the SQL Resource Provider accepts it, but an error is returned after the provider tries to process the FQDN. Resolution: FQDNs are now fully supported by the Resource Provider. When a tenant creates a VM, the name that is provided is not honored. Instead, a name is generated. Symptom: When tenants build a VM on premises, they expect the name that they provided to be honored as the NetBIOS name of the computer. This is not the case; the VM extension generates a name to maintain uniqueness in System Center Virtual Machine Manager. Resolution: Now the VM Resource Provider generates a name that’s based on the name that the user provides in the following pattern:Computer Name = name + “###” The numbers are added to the end of the name to guarantee uniqueness in System Center Virtual Machine Manager. The metrics that are associated with websites that are located in separate clouds whose endpoint credentials differ from each other are sometimes not retrieved; instead an error message is displayed. Symptom: When the tenant or administrator accesses information about websites that are located in clouds that are configured to use different endpoint access credentials, you receive the following error message from the Administration site:The server could not retrieve metrics (Internal Server Error). Resolution: The monitoring service responsible for retrieving the websites data now returns the correct metrics instead of an error. The SQL Resource Provider grants DB Owner permissions to tenants when during database provisioning. Symptom: When a SQL Server Database is provisioned by a tenant, the SQL Resource Provider grants DB Owner permissions to the corresponding SQL login. Resolution: Tenant logins that are created after you apply Update Rollup 3 are no longer granted DB Owner permissions. Existing tenant logins and already existing databases are not affected. To retroactively give existing tenant logins the same set of permissions and to remove the DB Owner level access, see the following TechNet article: Using SQL Server or MySQL with Windows Azure Pack ( http://go.microsoft.com/?LinkId=9851447 ) SQL Server Integrated Security or Windows Authentication is not supported by the Windows Azure Pack Tenant Site SQL extension. Symptom: When the tenant tries to create a database, no option for Windows Authentication is provided. Resolution: The SQL Tenant Site Extension now supports Windows Authentication. -> http://support.microsoft.com/kb/2965416

Hallo Community, seit heute steht das UR3 für die folgenden System Center 2012 R2 Produkte zur Verfügung: Data Protection Manager Operations Manager Orchestrator & Service Management Automation Service Manager Virtual Machine Manager -> http://support.microsoft.com/kb/2965090 Data Protection Manager Beim DPM werden die folgenden fünf Probleme behoben: A backup of a mirrored SQL instance fails if the principal SQL instance that was first backed up is now the mirror. DPM console crashes while a recatalog or “mark as free” operation is performed on an imported tape. The MSDPM service crashes when protected data sources have long names. The DPMRA service crashes during replica creation when the database name on one of the SQL instances matches or is a substring of a SQL instance name that is hosted on the protected server. This update lets administrators configure the DPMRA port and select a nondefault port. Daneben stehen nach der Installation des UR3 folgende Features zur Verfügung: Scalable VM Backup Backup and consistency check windows Support for synthetic fiber channel-to-tape -> http://support.microsoft.com/kb/2966014 Operations Manager Beim SCOM werden die folgenden 16 Probleme behoben: A deadlock condition occurs when a database is connected after an outage. You may experience this issue may when one or more HealthServices services in the environment are listed as Unavailable after a database goes offline and then comes back online. The Desktop console crashes after exception TargetInvocationException occurs when the TilesContainer is updated. You may experience this issue after you leave the console open on a Dashboard view for a long time. The Password expiration monitor is fixed for logged events. To make troubleshooting easier, this fix adds more detail to Event IDs 7019 and 7020 when they occur. The Health service bounces because of high memory usage in the instance MonitoringHost: leak MOMModules!CMOMClusterResource::InitializeInstance . This issue may be seen as high memory usage if you examine monitoringhost.exe in Performance Monitor. Or, the Health service may restart every couple of days , depending on the load on the server. The Health service crashes in Windows HTTP Services (WinHTTP) if the RunAs account is not read correctly. Windows PowerShell stops working with System.Management.Automation.PSSnapInReader.ReadEnginePSSnapIns. You may see this issue as Event ID 22400 together with a description of “Failed to run the Powershell script.” The PropertyValue column in the contextual details widget is unreadable in smaller widget sizes because the PropertyName column uses too much space. The update threshold for monitor “Health Service Handle Count Threshold” is reset to 30,000. You can see this issue in the environment, and the Health Service Handle Count Threshold monitor is listed in the critical state. An acknowledgement (ACK) is delayed by write collisions in MS queue when lots of data is sent from 1,000 agents. The execution of the Export-SCOMEffectiveMonitoringConfiguration cmdlet fails with the error “Subquery returned more than 1 value.” The MOMScriptAPI.ReturnItems method can be slow because a process race condition may occur when many items are returned, and the method may take two seconds between items. Scripts may run slowly in the System Center Operations Manager environment. When you are in the console and click Authoring , click Management Pack , click Objects , and then click Attributes to perform a Find operation, the Find operations seems unexpectedly slow. Additionally, the Momcache.mdb file grows very large. A delta synchronization times out on SQL operations with Event ID 29181. Operations Manager grooms out the alert history before an alert is closed. The time-zone settings are not added to a subscription when non-English display languages are set. Additionally, time stamps on alert notifications are inaccurate for the time zone. Web Browser widget requires the protocol (http or https) to be included in the URL. You cannot access MonitoringHost’s TemporaryStoragePath within the PowerShell Module. The TopNEntitiesByPerfGet stored procedure may cause an Operations Manager dashboard performance issue. This issue may occur when a dashboard is run together with multiple widgets. Additionally, you may receive the following error message after a time-out occurs:[Error] :DataProviderCommandMethod.Invoke{dataprovidercommandmethod_cs370}( 000000000371AA78 ) An unknown exception was caught during invocation and will be re-wrapped in a DataAccessException. System.TimeoutException: The operation has timed out. at Microsoft.EnterpriseManagement.Monitoring.DataProviders.RetryCommandExecutionStrategy.Invoke(IDataProviderCommandMethodInvoker invoker) at Microsoft.EnterpriseManagement.Presentation.DataAccess.DataProviderCommandMethod.Invoke(CoreDataGateway gateWay, DataCommand command) Da einige manuelle Schritte nach der Installation des UR3 erforderlich sind, unbedingt die Installation Notes und Installation Information beachten. Zu den manuellen Schritten zählen das Importieren von Management Packs und das Ausführen eines SQL Scripts auf der OperationsManager und der OperationsManagerDW Datenbank. -> http://support.microsoft.com/kb/2965445 Orchestrator & Service Management Automation Beim Orchestrator und der SMA werden die folgenden 10 und Probleme behoben: When a runbook calls other runbooks, Service Management Automation (SMA) concatenates all the participating runbooks into a single script and then passes the script to the PowerShell Workflow engine for compilation. The resulting script may contain multiple signature blocks, and SMA receives a compilation error from the Powershell Workflow. When child runbooks contain a signature block, the child runbooks cannot be compiled into inline runbooks. If there is more than one signature block, the resulting runbook will be corrupted. Cmdlets should request information from the server, up to a limit on the number of records that are returned, and then request the next “page” of records, until all records are retrieved. When a Windows PowerShell user uses the Set-SmaCertificate cmdlet, the user cannot receive private key information when he or she uses the Get-AutomationCertificate activity. The Automation menu stops working in Administrator Portal, and you may notice that the Orchestrator ODATA API controller cannot return Modules data. Additionally, you receive the following error message:Microsoft.Data.OData.OdataInnerError Message: The file exists. TypeName: System.IO.IOException Service Management Automation (SMA) resources are paged. The cmdlets have to evaluate the response from the web service for the presence of a continuation token and then make a sequence of ListNext web requests if a continuation token is present. SMA may not return the latest job ID for the runbook. The Invoke-ConfiguratorRunbook activity will have the job ID when it calls the Start-SMARunbook cmdlet, and the job ID should be returned to the caller. This occurs because SMA does not return the most recent job for a runbook. This issue occurs occur in a multiple-run scenario. If the configurator passes the first run, this issue has no effect. The Orchestrator Runbook Service (RunbookService.exe) may crash, and you receive the following error message:Application: RunbookService.exe Framework Version: v4.0.30319 Description: The process was terminated due to an unhandled exception. Exception Info: System.ServiceModel.CommunicationException If a runbook is large enough to exceed the default Windows Communication Foundation (WCF) channel size, an exception occurs. This issue causes the job to be stuck in the “New” state. When Update Rollup 2 is uninstalled, the connection properties to the instance of Microsoft SQL Server are cleared. -> http://support.microsoft.com/kb/2931268 Service Manager Beim SCSM werden die folgenden acht Probleme behoben: The Microsoft Systems Center 2012 Operations Manager configuration item (CI) connector does not retrieve disk mount point information for the Service Manager database. When a service request is created from a request offering in the console and when every user prompt is of the Simple List prompt type, the console shows options from all simple lists in each user prompt. When you try to close a customized change request, you cannot close the change request, the task throws an error, and the close is not actioned. MPSync job failures that are logged to the event log do not contain enough information to allow for a quick diagnosis of the problem. All Service Manager workflows may stall when any invalid XML characters (control characters) are used in a work item property. Using the “Set First Response or Comment” task in a service request creates a “Private” flag that cannot be used for evaluation in a notification action log or a user comment update. This behavior occurs because the private flag is undefined. If the configuration management database (CMDB) is down (that is, if it is offline or unable to connect because of network issues) at the time that the Health service is restarted and if the CMDB continues to be offline for a long time, the Health service may enter an unrecoverable state even if the database comes online again later. Configuration Manager Connector synchronization may stop after Update Rollup 2 for Service Manager 2012 SP1 or Update Rollup 6 for Service Manager 2012 R2 is applied. This issue occurs when the connector tries to synchronize malformed or incomplete software version information from the Configuration Manager database. -> http://support.microsoft.com/kb/2962041 Virtual Machine Manager Beim VMM werden die folgenden 31 Probleme behoben: Total storage for a User role is reported incorrectly. For example, the User role can use only half of the allowed quota. A host cluster update fails intermittently because of a locked job. Virtual machine (VM) refreshers do not update highly available virtual machines (HAVMs) after failover to another node. A cluster IP address for a guest cluster configuration in a Hyper-V Network Virtualization (HNV) environment is not updated correctly by using HNV policies during failover. For more information about this issue, see the following article in the Microsoft Knowledge Base: 2981736 The cluster IP address in an HNV environment is updated incorrectly during failover Server Message Block (SMB) shares may not be usable by high availability (HA) workloads if they are also connected to stand-alone hosts. Storage objects discovery does not occur on a Virtual Machine Manager server if the discovery item is too big. A Virtual Machine Manager job that assigns network service backend connectivity fails. Enable maintenance mode fails when you evacuate failed-state VMs. The Virtual Machine Manager service cannot be restarted because of database corruption. The ZH-TW language incorrectly appears in the tooltip of the VM Network icon. Library refresher rewrites the alternative data stream on every file during every update. For iSCSI hardware storage-based array, when the MaskingPortsPerView property option is set to “multi-port per view,” the target endpoint is not obtained as the port address. The virtual hard disk (VHD) is left in the source folder after storage migration is completed. The addition of a bandwidth limitation to an existing virtual private network (VPN) connection is not added to the generated script. A VM that is attached to an HNV VM network loses connectivity when it is live migrated to another node in the failover cluster that is not currently hosting other VMs that belong to the same VM network. VM network shared access is lost after a service restart or an update interval. The Remove-SCFileShare command fails for a network-attached storage SMI-S provider. Setting the template time zone to UTC (GMT +0:00) is incorrectly displayed as “Magadan Standard Time.” The System Center 2012 R2 Virtual Machine Manager crashes when you add groups that contain the at sign (@) character in User roles. VM deployment fails in a VMWare environment when you have virtual hard disk (.vmdk) files of the same size in your template. Deployment of an application host on HAVMM fails and generates a 22570 error. Live migration of an HAVM cross cluster creates a duplicate VM role in the target cluster. An error occurs when you apply physical adapter network settings to a teamed adapter. A VMM agent crashes continuously when the HvStats_HyperVHypervisorLogicalProcessor query returns a null value. A host refresh does not update the VMHostGroup value of a VMWARE cluster after the cluster is moved from VCENTER. VMM reports an incorrect Disk Allocation size for dynamic VHDs that are mapped to a virtual machine. A VMM service template script application does not work for a self-service role. VM creation fails if Virtual Machine Manager undergoes failover during the creation process. The Access Status value of a file share is incorrect in the user interface. The Virtual Machine Manager service crashes because of an invalid ClusterFlags value. VMs cannot be deployed from a service template to a cloud across multiple host clusters (multiple datacenters). Daneben stehen nach der Installation des UR3 folgende Features zur Verfügung: Ubuntu Linux 14.04 (32-bit) Support Ubuntu Linux 14.04 (64-bit) Support Host DHCP extension driver upgrade Several performance improvements Several Management Pack package improvements Nach dem Update muss ein SQL Script ausgeführt werden, daher gilt auch für den VMM unbedingt die Installation Instructions beachten. -> http://support.microsoft.com/kb/2965414

Hi, Generation 2 = UEFI Generation 1 = BIOS VMware Workstation verwendet standardmäßig das normale BIOS. UEFI muss explizit aktiviert werden. -> https://community.landesk.com/support/docs/DOC-27736

Hi, sind der Desktop und die VM auf dem aktuellen Patchlevel?

Hi, Geschmackssache sage ich jetzt mal oder wie es so schön heisst es kommt darauf an. Wenn ich den SCVMM nutze, dann würde ich keine Sicherung des Hyper-V Hosts machen, sondern die Zeit die ich für die Test für den DR Fall brauche in das Bare-Metal Deployment stecken. Damit kann ich den Hyper-V Host vollkonfiguriert innerhalb einer Stunde wieder voll in Betrieb nehmen. Aber wie gesagt es kommt darauf an. Ich zum Beispiel sichere die Hyper-V Hosts nicht und auch sonst kann ich nicht wirklich einen Trend dahin gehend erkennen. Just my two cents.

Hi, in dem Blogartikel wird nur beschrieben wie man das ARP Spoofing unterbindet. Standardmäßig ist dies nämlich im vSwitch erlaubt. Siehe hier: http://msandbu.wordpress.com/2013/04/03/arp-guard-in-hyper-v-2012/. Ansonsten zusätzlich mal das MAC Spoofing für die vNIC der VM erlauben.

Ich habe mal die Umfrage entfernt, die hatte ja mit dem Thema nichts zu tun. ;)

Hi, auch wenn Windows Server 2012 R2 nicht unterstützt wird. Installiere, falls noch nicht getan, den folgenden Hotfix für Windows Server 2008 R2 Hyper-V. Damit laufen Windows Server 2012 VMs erst stabil. Eventuell hilft es auch in deinem Fall. -> http://support.microsoft.com/kb/2744129

Danke für die Rückmeldung! :)

Hi, zu deinen beiden Fragen, schau dir mal folgende Links an: -> http://social.technet.microsoft.com/Forums/systemcenter/en-US/1577d5c8-9929-46c0-b6eb-cf55944c293e/scom-2007-r2-agent-what-processes-are-running -> http://technet.microsoft.com/en-us/library/hh212808.aspx Ansonsten den SCOM mal installieren und schauen welche Prozesse/Dienste hinzukommen. Kann leider derzeit nicht auf meine Testumgebung zugreifen.

Habe das Thema mal ins richtige Subforum verschoben. :)

Gerne! Ist aber ein bekanntes Problem Broadcom NICs und VMQ. ;)

Hast du die Treiber vom Hersteller installiert oder die Standardtreiber von Windows? Eigentlich sollte diese Einstellung in den NetXtreme NICs zu finden sein, wenn der Server nicht allzu alt ist.

Hi, welche Netzwerkkarten stecken in dem Server und ist VMQ auf den NICs aktiviert?

Hi Norbert, beim Exchange kann ich dir es nicht sagen. Aber im oben genannten Link steht auch für Hyper-V zum Beispiel Supported, but not recommended, daher ist es empfehlenswert das CNO zu nutzen. Ohne Live Migration macht nämlich ein Hyper-V Cluster aus meiner Sicht nicht wirklich viel Sinn. Man verliert ja Funktionalitäten, wenn man darauf verzichten kann schön und gut. Dann sollte man aber später nicht meckern, wenn irgendetwas nicht funktioniert. ;)

@Doso Das was du meinst ist, dass man für den Failover Cluster nicht mehr zwingend das Cluster Name Object im AD benötigt, sondern nur einen DNS Eintrag. Aber es müssen immer noch beide Server Mitglied der Domain sein. Empfehlenswert ist es aber das CNO zu nutzen.

Hi, kannst du ohne Probleme machen. Sobald man mit dem einem Hyper-V Server und einem weiteren einen Cluster bilden möchte, müssen beide Mitglied einer Domain sein. ;)

Hi, wie Doso es beschrieben hat, ist es der korrekte Weg. Eine Alternative bei vielen VMs sind die VMM PowerShell Cmdlets.