Replies posted by Nupur.Sakhalkar

46 Replies

asked in AHV Virtualization

Alert PE-PC Connection Failure

@Welton kindly refer to KB-6970 to resolve PE-PC Connection Failure alerts seen in the cluster

Pablo AlejandroVoyager

asked in Installation & Configuration

LCM Not Showing AOS Upgrades

@Kcmount Yes, I can understand. Unfortunately, it was a design decision made to avoid suggesting an upgrade option to an AOS version that is already EOL. Nutanix ENG team is actively working on implementing INFO messages that can be shown on the LCM page to inform about AOS versions going EOL beforehand to avoid this situation. For now, LCM should still offer an upgrade option in scenarios wherein customers can upgrade from EOL AOS to a Supported AOS release version.However, in situations like the above wherein the upgrade path is to upgrade from one EOL AOS version to the another AOS version that is also EOL, in this case, both LCM and 1-click won’t show these upgrades since the target version is EOL. In this case, the only option would be to follow the old procedure and manually upload the AOS binary and metadata files in prism to upgrade AOS in the cluster.

Pablo AlejandroVoyager

asked in Installation & Configuration

LCM Not Showing AOS Upgrades

@Pablo Alejandro LCM won’t show any AOS updates already at the end of Support Life (EOL). As per the upgrade path guide (considering LTS releases only), from AOS - 5.10, you cannot directly upgrade to the latest AOS LTS release. Instead, it will be a 2-path upgrade process:AOS 5.10 => 5.15.7 LTS AOS 5.15.7 LTS => 5.20.4.5 LTS However, the first path upgrade to AOS 5.15.7 LTS version won’t be shown in LCM since AOS 5.15.7 AOS went EOL in May 2022, as mentioned on the EOL Information page on Portal. This is why you cannot see any AOS upgrades on Portal. For the same reason, this AOS 5.15.7 version won’t appear as a 1-click download available on the ‘Upgrade Software’ settings page on Prism. Hence, in this case, the only way to upgrade AOS to the 5.15.7 version is to upload the metadata and binary files on the ‘Upgrade Software’ settings page on Prism for AOS, and once this is done, you can proceed with upgrading AOS on the cluster. Once AOS is at 5.15.7, you can re-run LCM invento

chenzh4Trailblazer

asked in How It works

inter_cvm_connections check in NCC 4.X

@chenzh4 The ping_all.INFO file will only show up ping latencies above 0.5 ms which is the threshold for this one. The latency value which is < 0.5 ms threshold won’t be shown as that is an expected value range. This is why you could only see the entries for gw since the ping latencies between CVM and gw are > 0.5 ms. In case your cluster sees high ping latencies between CVMs intermittently, then those CVM-CVM or CVM-host ping latencies should be populated around the issue time in this file since at that time, I will expect the ping latency value to go > 0.5 ms.

chenzh4Trailblazer

asked in How It works

inter_cvm_connections check in NCC 4.X

@chenzh4 There should be a ping_all.INFO file present inside /home/nutanix/data/logs/sysstats/ directory on each CVM which should give you the ping response times between CVMs and CVMs-hosts.

AngeTrailblazer

asked in Nutanix Disaster Recovery

Federation on mutilple Prism central Instances

@Ange Unfortunately, any Prism Element (PE) cluster can only be registered to one Prism Central (PC). Considering that you are planning to use Leap On-Prem for 2 PE clusters which are installed on different datacenter locations, hence two Prism Centrals (one PE cluster will be registered to one PC respectively) will be needed, each deployed and registered to their respective PE cluster.Since the PE clusters are already be registered to their individual PC instances, hence there cannot be any third PC instance that can be registered to these PE clusters.

daniel.martinez-42326Adventurer

asked in Intelligent Operations

How to remove several VM at the same time on AHV cluster?

@pkjarora You can use this command instead to forcefully say yes for deleting each VM listed in the /tmp/vms_to_delete file instead of it asking for confirmation on deleting each VM:for i in $(cat /tmp/vms_to_delete); do echo $i && (echo yes | acli vm.delete $i); done

jfm_esiPathfinder

asked in Installation & Configuration

AOS 5.20.2 shown, can't upgrade to 5.20.3, says 5.20.3 already there!

@jfm_esi In order to further investigate this behavior, kindly answer the following questions:Did you run LCM inventory before performing the AOS upgrade to 5.20.3? If yes, then did the LCM inventory complete successfully? If LCM inventory was not run before the AOS upgrade, then can you perform LCM inventory and then check if LCM shows the AOS version to be 5.20.2 or 5.20.3 in this case. What current AOS version is shown when you log in to PE web console > Settings Dashboard > Upgrade Software > AOS? Is it showing the 5.20.2 or 5.20.3 version?If your cluster is actually running on AOS 5.20.3 version, however, an LCM inventory was not run before the upgrade, then it means that LCM was not updated with the AOS current version details 5.20.3 and hence the LCM inventory might not be reflecting the correct current AOS version.

neyrignouxAdventurer

asked in Installation & Configuration

Which vCenter 7 version recommended for AOS 5.20 LTS

@neyrignoux AOS version doesn’t have any compatibility matrix with vCenter directly. However, we do have a compatibility matrix between AOS and Hypervisor versions (ESXi in your case). As per the Compatibility Matrix available in Support Portal, AOS 5.20.3 is compatible with the following ESXi 7.x hypervisor versions:ESXi 7.0 U2a ESXi 7.0 U2 ESXi 7.0 U1 ESXi 7.0So, as long as you use a vCenter 7.x version that is compatible with any one of the above ESXi 7.x versions, that should work without issues with AOS 5.20.3 version. Kindly let me know if you have any more questions on this one.

BrianDonnanPathfinder

asked in How It works

HDD/SSD Configuration for 3-Node Chassis

@BrianDonnan Kindly refer to the following documentation for NX-3060-G5 system specifications - https://portal.nutanix.com/page/documents/details?targetId=System-Specs-G5-Multinode:sys-system-specs-nx3060g5-r.htmlAlso, as mentioned by @JamesDyke earlier, as per the naming terminology for NX Hardware https://portal.nutanix.com/page/documents/details?targetId=System-Specs-G5-Multinode:System-Specs-G5-Multinode, the naming convention for NX-3060-G5 HW model is NX-3X60-G5 where X is the number of nodes present in the block. If your case, since you have 3 nodes, it will be NX-3360-G5Also, the only difference between 3-nodes and 4-nodes in the NX-3060-G5 HW model is that the last node slot (located in position D) should be empty. Kindly refer to the NX-3060-G5 Node Naming guide for more details on this one. Regarding the replacement model number for this node, it will be NX-3060-G5 hardware model node. If this Nutanix asset has a valid support contract, then it will be available for replace

Arun KumarVoyager

asked in AHV Virtualization

Foundation and Phoenix

@Arun Kumar Glad to help you out on this one!

Arun KumarVoyager

asked in AHV Virtualization

Foundation and Phoenix

@Arun Kumar Here are the answers to your questions about the phoenix/foundation and LCM based upgrades:1. Once the phoenix iso is created and mounted to the host along with the upgrade bundle, Can phoenix run independently and complete the upgrades or needs help from foundation for performing some tasks like host reboot?--- Phoenix process cannot run independently during firmware upgrades. It still needs help from the foundation service which runs on the LCM leader CVM that will control all the other tasks like rebooting node to phoenix and then after upgrade, rebooting the node back to phoenix, etc. 2. During LCM upgrade, since the node undergoing upgrades, will have all VMs powered off including the CVM, will it take help from a remote CVM foundation to orchestrate the upgrades on this host?--- During LCM upgrade, the host goes into maintenance mode. So, all the user VMs (except the pinned VMs) migrate off to another hosts in the cluster and the CVM which resides on this host is pow

AbdulrhmanTrailblazer

asked in Installation & Configuration

Warning Message After Upgrading AHV

@Abdulrhman The above WARN alert message “WARN: AHV version 20190916.360 is currently installed on host x.x.x.10, while the minimum compatible AHV version is 20201105.2244” indicates that the AHV version on host x.x.x.10 is at 20190916.360 version.On Nutanix AHV clusters, there is a compatibility between AOS and AHV versions. Since the AOS version is at 5.20.2.1, the following AHV version is compatible with it - AHV-20201105.2244. Now, this doesn’t mean that the cluster will stop working if the current AHV version (20190916.360) doesn’t match the AHV version compatible with AOS - 5.20.2.1 (AHV-20201105.2244). It will continue to work; however, it is recommended to upgrade the AHV version to the version compatible with the current AOS version on the cluster. Since the AOS version is 5.20.2.1, it recommends upgrading AHV to the 20201105.2244 version. One way to check the current AHV version on all nodes present in the cluster is to run the following commands from any one CVM (login as nu

MaheshNikamPathfinder

asked in How It works

I want to update VM only with Categories, but asking for spec_version

@MaheshNikam To get the spec_version and entity_version, you will need to first use the GET REST API call for VM (https://AddressOfServer:9440/api/nutanix/v3/vms/{VM-UUID} ), and once you get those details, you can then use it in the above-mentioned PUT API call for updating VMs via REST API.

Kishore SalipalliExplorer

asked in Installation & Configuration

Offline foundation applet not detecting nodes (Lenovo)

@Kishore Salipalli The CVM IPs which are in the 192.168.5.x subnet is used only for CVM-hypervisor communication. Those 192.168.5.2 CVM IPs are not used for external CVM communication. Each CVM and AHV hypervisor has two networks associated with it:External network - which is used to connect to the AHV host/CVM externally (which is 172.31.199.x network in your case) Internal Nutanix network - generally used for CVM-host storage communication (192.168.5.x network). This network is not recommended to be used for external CVM/host communication. This internal network is set up during the initial node imaging and should not be modified later. Hence, in your case, you will need to configure external IP addresses for CVMs as well. (belonging to the same subnet as AHV hosts - 172.31.199.x network). Since you have not created a cluster as of yet and only have the factory imaged nodes (i.e., AHV hypervisor + CVM is present on the node), hence you can use the following guide to configure CVM ext

Jean-baptisteWayfarer

asked in Installation & Configuration

Unable to create cluster - New install

@Jean-baptiste Foundation is failing with error - “Failed to validate AOS package lcm_nos_5.20.1.1.tar.gz” because it looks like wrong AOS bundle was uploaded in foundation.Based on the above error, AOS LCM Bundle - LTS (Version: 5.20.1.1) with filename: lcm_nos_5.20.1.1.tar.gz is being used instead of using the nutanix_installer_package-version#.tar.gz file for AOS install via foundation. Hence in order to resolve this issue, kindly go to AOS Downloads Page on the Nutanix Support portal and download the nutanix_installer_package-version#.tar.gz file for AOS 5.20.1.1 (filename - nutanix_installer_package-release-euphrates-5.20.1.1-stable-726ea8f7dc4bca156d3e3f63cd7982eecb70c8cb-x86_64.tar.gz) which will be used for imaging the nodes with desired AOS release via foundation. Kindly refer to the Foundation Field Installation Guide for more details on this one.

NoDowtAdventurer

asked in Installation & Configuration

License Expiry Question

Yes, I can understand! Glad to help you out on this one.

Ivan_lopezExplorer

asked in Installation & Configuration

Expand 1 node cluster

@Ivan_lopez No, a single-node or two-node cluster cannot be expanded later on. So if you have a single node cluster currently, it cannot be expanded to a 2 node or 3+ node cluster. Kindly refer to KB-5943 for more details on it. Cluster expansion is only allowed for clusters which had 3 or more than 3 nodes in the cluster when the cluster was initially created.

NoDowtAdventurer

asked in Installation & Configuration

License Expiry Question

@NoDowt There should not be any impact to your cluster. License expiry will generate a red banner on top of Prism page stating some details about an expired license. However, having an expired license on Prism element cluster doesn’t affect the cluster performance and functionality as of today.

AtherVoyager

asked in Installation & Configuration

Upgrading sequence AOS, NCC, Foundation

@Ather You can directly upgrade foundation to 5.0.4 and NCC to 4.2.0.2 version. No need to upgrade it to middle older versions for these two components.

GOKAdventurer

asked in Intelligent Operations

Filter UVMs that need an NGT upgrade

@GOK FEAT-10795 is created on our end to filter out VM list which needs NGT upgrade via Prism Central.

GOKAdventurer

asked in Intelligent Operations

Filter UVMs that need an NGT upgrade

Unfortunately we currently don’t have search/filter option available to only display VMs which needs NGT upgrade on the cluster. We already have a FEAT request created for this one on our end but there is no ETA currently as to when will this feature be available on Prism Central. In the meantime, you can:Select the option to only show VMs which has NGT installed in it Once the VM list is filtered out to only display VMs that has NGT installed on it, then sort the “Installed version” column by clicking on the “Installed version” column name to display the VMs which need NGT upgrade on top of the NGT installed VMs filtered list. The above workaround doesn’t exactly give only the VMs which needs NGT upgrade but it will at least simplify the task of finding such VMs.

MentirosoExplorer

asked in How It works

How does Prism display memory installed?

The host memory capacity displayed by Prism is the memTotal value which can be used by the AHV hypervisor. This value can be obtained by looking into the MemTotal value listed in the cat /proc/meminfo command output on the AHV host:cat /proc/meminfo | grep MemTotalNow even though the MemTotal value in this file is visually mentioned as “kB”, however it is actually in kibibytes as per the Redhat documentation - https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/6/html/deployment_guide/s2-proc-meminfo So, you can convert this kibibytes values to Gigibytes and this value should match with what is shown under host’s memory capacity in Prism web console The same memory capacity value can also be found using free -g command output (shows memory values in GiB) on the AHV host and looking into the Mem Total value in it.free -gSince Mem Total value is directly shown in GiB in the above output, you don’t need to convert it to GiB again - this value should match to what is bei

daniel.martinez-42326Adventurer

asked in Installation & Configuration

Minimun VM witness requested

@daniel.martinez-42326 There are differences in the details mentioned in the above two links because the next community link (https://next.nutanix.com/installation-configuration-23/metro-availability-witness-option-38374) is for an older AOS version. For newer AOS releases (AOS 5.15.x and above), One Witness VM can support up to 100 instances, that is the combined total number of protection domains in the registered Metro cluster pairs and the number of registered two-node clusters. (A single Witness VM can support both metro availability and two-node clusters.) Since the portal link shared by you is for AOS 5.18, hence it states about this limit in it.So if your AOS version is 5.15.x or above, then in this case, a single Witness VM can support up to 100 instances - this value is the combined total number of protection domains in the registered Metro cluster pairs and the number of registered two-node clusters. If the total number of 2-node clusters is 70, then you can use a single Wit

LinuxInsideVoyager

asked in How It works

ISO imported into Images have no size can't mount as CD

@LinuxInside Unless the image which you upload to Prism is in active state, it cannot be used as CD-ROM for any VMs. The primary reason as to why an image might not be in active state (basically no size seen and image not active) is because the image uploaded to Prism was corrupt/incomplete. This usually happens when you are uploading larger file size images via your local computer using any web browser. Most of the internet browsers have some upper limit in the file size that can be uploaded to it and so the image which gets uploaded to Prism might be corrupted/incomplete which in turn causes it to stay inactive on the cluster. In order to avoid such scenarios, if the ISO file size or the disk image size is large (in GBs), then it is recommended to use the URL method to upload the image to the cluster instead of uploading the image via local computer since in this case it should not depend on the internet browser limits or any minor disruptions that can happen for local browsers. Kind

1
2

Page 1 / 2