5 Essential Tips for Maximizing Your Experience at Nutanix .NEXT for Bloggers
When you say "made Prism fail", does it mean you can no longer access the Web Console? If so, please open a support case. Otherwise, could you provide us with a screenshot of the upgrade window?We do not have phusion passenger on our toolchain. Thank you
Hi , For the official statement about 3rd party components, please visit the following URL: [url=http://www.nutanix.com/support/support-faqs/#non-nutanix-qualified-3rd-party-components]http://www.nutanix.com/support/support-faqs/#non-nutanix-qualified-3rd-party-components[/url] "If a customer uses a third-party component in a Nutanix device, and a fault is traced to the use of this third-party component, then at the discretion of Nutanix, support and warranty service may be withheld."
It sounds like a mismatched SSL issue. Did you install any SSL certs be fore the upgrade? If you have support, please open a case, mentioning that it **may** be similar to ENG-29969
Can you confirm which build of ESXi 5.5 was being used. It needs to be a minimum of the build starting with 245
Are the times between your cluster CVMs and the desktop you're using more than 5 minutes apart?
As you said, there may be stuck tasks in Curator. Please open a support case, as some diagnosis needs to be done as there may be a couple paths to take to solve this. Thanks
If the cluster is up & running (if there are enough healthy CVMs), run the command: ncli cluster info It will return the cluster ID, i.e: ncli cluster info Cluster ID : 1212 On a 4.x cluster: ncli cluster info Cluster ID : 0005119d-9bc8-5117-0000-000000000272::626 In this case the number after the "::" is your cluster ID If the cluster is not running, check the other CVM's factory config file, the lower cluster ID is the one to use.
In foundation 2.0, these are the supported hypervisors: ACCEPTED_HYPERVISOR_ISOS = {"fa6a00a3f0dd0cd1a677f69a236611e2": ("esx", "5.0u2"),"391496b995db6d0cf27f0cf79927eca6": ("esx", "5.0u3"),"2cd15e433aaacc7638c706e013dd673a": ("esx", "5.1u1"),"6730d6085466c513c04e74a2c2e59dc8": ("esx", "5.1u2"),"9aaa9e0daa424a7021c7dc13db7b9409": ("esx", "5.5"),"5d109d10864ea8ad1d186a336bd26c1c": ("esx", "5.5u2"),"fb101ed6d7328aca6473158006630a9d": ("hyperv", # From MSDN.("2012r2", ["datacenter", "standard"])),"b52450dd5ba8007e2934f5c6e6eda0ce": ("hyperv", # From EA portal.("2012r2", ["datacenter", "standard"])),"9c9e0d82cb6301a4b88fd2f4c35caf80": ("hyperv", # Hyper-V free.("2012r2", ["free"])),"b52450dd5ba8007e2934f5c6e6eda0ce": ("hyperv", # 2012r2 with update from MSDN.("2012r2", ["datacenter", "standard"])),} In foundation 2.0.2, these are the supported hypervisors: "fa6a00a3f0dd0cd1a677f69a236611e2": ("esx", "5.0u2"),"391496b995db6d0cf27f0cf79927eca6": ("esx", "5.0u3"),"2cd15e433aaacc7638c706e013
If I understand correctly, you are adding a new node to an existing cluster? Has the node being imaged with the same hypervisor? What about NOS version? What error is returned when you try to add the node?
Failover clustering is not supported on NOS with ESXi due to the lack of support of RDM's and SCSI 3 persistent reservations.This is targeted to be addressed in a not so far future release. Theoretically it should be possible with shared VHD on SMB3 with Hyper-V, but no testing has been done on that (AFAIK). SQL Server Always On and other non-shared disk clustering techniques are the only way of doing clustering on NOS with ESXi.At a stretch you could create a Windows VM as an iSCSI target and use that to form the cluster. Not an elegant solution however.
merter, Was a case opened for this yesterday? I vaguely remember another SRE mention this scenario...
Thanks, I see now what you mean; I personally have seen only one deployment using more than a couple logical networks, but if planned correctly, it should work fine. Regarding the CVM traffic (I/O) goes over the Internal switch unless the local CVM has a problem, in that case traffic is re-routed over to another CVM, so host Management and CVM eth0 would fall under the same VLAN
Our commitment is to complete qualification testing < 90 days after GA; and I'm glad to report we're on track on that commitment Cheers
David, I do not have screenshots at this time, but you're correct on your assumptions about the 10GigE, and the 1GigE adapters are usually disconnected after the setup has been completed: PS C:Usersf3rm1n> Get-NetLbfoTeam Name : NetAdapterTeamMembers : {Ethernet 2, Ethernet 3}TeamNics : NetAdapterTeamTeamingMode : SwitchIndependentLoadBalancingAlgorithm : Dynamic PS C:Usersf3rm1n> Get-NetLbfoTeamMember Name : Ethernet 2InterfaceDescription : Intel(R) 82599 10 Gigabit Dual Port Network ConnectionTeam : NetAdapterTeamAdministrativeMode : ActiveOperationalStatus : ActiveTransmitLinkSpeed(Gbps) : 10ReceiveLinkSpeed(Gbps) : 10FailureReason : NoFailure Name : Ethernet 3InterfaceDescription : Intel(R) 82599 10 Gigabit Dual Port Network Connection #2Team : NetAdapterTeamAdministrativeMode : ActiveOperationalStatus : ActiveTransmitLinkSpeed(Gbps) : 10ReceiveLinkSpeed(Gbps) : 10FailureReason : NoFailure
You add the hosts manually to the domain (sconfig); after that, expand the cluster from Prism, then setup_hosts (won't reboot as they're already in the domain), then setup_scvmm.
In NOS 4.1.1.4, zookeeper should enforce the CVM to acquire the timezone when a new node is added. I am not sure it does not in previous versions, so a cluster could be running with CVMs in different timezones, but as long as the epoch time matches there is no problem in functionality. NCC will display a Warning if this condition is found.
We have an improvement in this area that is currently targeted for 4.1.3. For now, we can provide a workaround. Please open a support case and mention you need a custom refresh time for Prism.
You can change the timezone withthe following command: ncli cluster set-timezone timezone=cluster_timezone**Replace cluster_timezone with the timezone of the cluster, timezones are defined in /usr/share/zoneinfo (for example, America/Los_Angeles, Europe/London, Asia/Tokyo).
I know this is not for Symantec Backup Exec, but take a look at: http://ntnx-portal.s3.amazonaws.com/solutions/Best%20Practices/Veeam%20on%20Nutanix%20with%20VMware%20Best%20Practices.pdf It may give you some ideas as to how to implement it.
I have no more images to share; I don't know if Symantec caould provide an agent to be installed on a proxy server that would get the backup data from the agents in each VM and store it on tape, again, it would be a configuration like the image above. I believe there is also a BackupExec appliance that could perhaps accomplish this. Cheers
What is the OS of the VM and what type of scsi controller is it using for the virtual disks? I would suggest to add a second SCSI controller to the VM and have the E: drive connected through it, Use the VMware Paravirtual controller as the type for both
Hi saquib382, I assume the purpuse of this tape vault os for backup of your VMs/data. Are you using an specific backup utility like Veeam? Since our nodes do not have a FibreChannel HBA, I would suggest the use of a proxy server that would be connected to the Tape Vault. You would use the vmkernel management network (10GigE) to connect the cluster to the proxy, something like the following image. [img]https://d1qy7qyune0vt1.cloudfront.net/nutanix-us/attachment/360i1684D0B219D92114.png[/img]
For existing KVM VMs, we have a simple process to introduce those legacy (pre NOS 4.1) VMs to Acropolis. Legacy VMs and Acropolis managed VMs show up in Prism, but we can only edit the Acropolis managed VMs in the Prism interface. To import the VM in to acropolis, the import_vm utility is used. It is not necessary to power off the VM. Check the following doc: https://portal.nutanix.com/#/page/docs/details?targetId=KVM_Administration-NOS_v4_1:man_import_vm_r.html
Can you please elaborate more on the issue and explain what is that you want to achieve? Do you want to vMotion the CVM, or connect one of its vNics to a different vSwitch? Cheers
I assume, from the day the op was created, that the diagnostics have run and finished... However, for future reference, the KB linked above recommends: 1) The first one is to verify that ESX is configured correctly to look up against the correct DNS server. 2) If resolution 1 is not possible, you can edit ssh_config file on the hypervisor to include the following line at the bottom: useDNS no
Already have an account? Login
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
Sorry, we're still checking this file's contents to make sure it's safe to download. Please try again in a few minutes.
Sorry, our virus scanner detected that this file isn't safe to download.
