NS0-593 Questions and Answers

The VMware feature that is supported by ONTAP software to identify and follow I/Os from each VM on the shared FC LUN-backed datastores is Virtual machine ID (VMID). VMID is a unique identifier assigned to each VM by the vCenter Server. ONTAP software can use VMID to map the VMs to the FC LUNs and provide visibility into the I/O performance and latency of each VM. This feature is also known as Virtual Storage Console (VSC) for VMware vSphere12. References: 1:Virtual Storage Console for VMware vSphere | NetApp Documentation 2: How to use Virtual Storage Console (VSC) to monitor the performance of virtual machines - NetApp Knowledge Base

 To create a packet trace for all clients on the 192.168.9.0/24 subnet, you have to run the packet trace on the interface that has an IP address in the same subnet. According to the exhibit, the only interface that matches this criterion is e2a-100, which has an IP address of 192.168.3.1/24. The other interfaces have IP addresses in different subnets, such as 10.92.4.202/16, 10.91.3.132/24, 10.95.7.114/24, and 10.92.6.200/16. Therefore, the correct answer is B. e2a-100. References = How to capture packet traces on ONTAP 9.10+ systems, How to capture packet traces (tcpdump) on ONTAP 9.2 to 9.9 systems, How to download captured packet traces from ONTAP

= The FIFO message before NMI is useful for investigating the issue because it shows the state of the system before the non-maskable interrupt (NMI) was triggered by the L2 watchdog timeout. The FIFO message contains information about the CPU registers, the stack pointer, the instruction pointer, and the last executed instructions. This can help identify the cause of the system hang or deadlock that led to the watchdog reset. The other FIFO messages are not useful because they show the state of the system after the reset or shutdown, which may not reflect the original problem. References = https://kb.netapp.com/onprem/ontap/hardware/Handling_L2_Watchdog_Resets_on_the_FAS8200_and_AFF_A300_platforms

https://docs.netapp.com/us-en/ontap-metrocluster/install-ip/task_sw_config_restore_defaults.html

= The error message indicates that the volume my_data_vol is WAFL inconsistent, which means that there is a discrepancy between the data blocks and the metadata in the file system. This can be caused by various factors, such as hardware failures, software bugs, power outages, or network disruptions. The first action to correct this issue is to determine the root cause behind the inconsistency before attempting any recovery procedure, as recommended by the NetApp documentation1. This is because some recovery procedures, such as wafliron or storage takeover, may not work or may cause further damage if the underlying cause is not resolved. For example, if the inconsistency is due to a faulty disk or shelf, running wafliron may not fix the problem and may even corrupt more data. Therefore, it is important to identify and isolate the cause of the inconsistency before taking any further steps. References = 1 Volume Showing WAFL Inconsistent - NetApp Knowledge Base

 The two throughput elements that are controlled by host applications are operation size and concurrency. Operation size refers to the amount of data that is transferred in a single I/O request, and concurrency refers to the number of I/O requests that are issued simultaneously. Both of these factors affect the throughput of a host application, as they determine how much data can be sent or received within a given time period. Larger operation sizes and higher concurrency can increase the throughput, but they may also increase the latency and resource consumption12. References: 1: NetApp Performance Analysis on Data ONTAP - Chapter 3: Host-Side Performance 2: NetApp Performance Analysis on Data ONTAP - Chapter 4: Storage System Performance

 = The Connector is a software component that enables Cloud Manager to manage resources and processes within your AWS environment. To deploy a Connector in AWS, you need to have the correct IAM policies attached to your user account or role. These policies grant the Connector the permissions to perform various actions on your behalf, such as creating and deleting EC2 instances, VPCs, security groups, and so on. If you do not have the required IAM policies, the deployment will fail with the message ‘Insufficient permissions to deploy Cloud Connector’. You can find the list of required IAM policies in the Connector deployment policy for AWS. References = Creating a Connector in AWS from Cloud Manager, page 3-4.

 The wafl.dir.size.max event occurs when a directory has reached its maximum directory size (maxdirsize) limit, which prevents new file creation in that directory. The maxdirsize is a volume-level option that can be modified using the volume modify command. To solve this issue, you should increase the maximum directory size for the volume that contains the CIFS share, as long as it does not exceed 3% of the physical memory. Alternatively, you can reorganize the directory structure to avoid having too many files in one directory. References =

• Max Directory size error - NetApp Knowledge Base
• How to identify the target directory of wafl.dir.size.warning - NetApp Knowledge Base
• Post about wafl.dir.size.max | Source.kohlerville.com
• WAFL Max dirsize | uadmin

• The question refers to a MetroCluster configuration, which is a disaster recovery solution that uses two physically separated, mirrored clusters1.
• The exhibit shows a MetroCluster switchover scenario, where Site A has experienced a disaster and Site B has taken over the tasks of Site A2.
• To return the MetroCluster to a normal state, you need to perform a MetroCluster switchback operation, which reverses the switchover and activates the original sync-source storage virtual machines (SVMs) on Site A3.
• To perform a MetroCluster switchback, you need to enter the metrocluster switchback command on the cluster that was the source of the switchover, which is Site A in this case3.
• The other options are not correct, because:
• Understanding MetroCluster data protection and disaster recovery - NetApp
• Perform IP MetroCluster switchover and switchback - NetApp
• Performing a switchback - NetApp
• High-availability configuration - NetApp

• NFSv4 is a network file system protocol that supports security, performance, and scalability features. NFSv4 uses ID mapping to ensure that the permissions of files and directories are consistent across different NFSv4 servers and clients1
• ID mapping is the process of translating the user and group identifiers (UIDs and GIDs) of the local system to the user and group names (user@domain and group@domain) of the remote system, and vice versa. ID mapping is done by the idmapd service, which uses the /etc/idmapd.conf file to determine the domain name of the system2
• ID mapping requires that the NFSv4 server and client have the same domain name configured in the /etc/idmapd.conf file. If the domain names do not match, the idmapd service cannot map the UIDs and GIDs to the user and group names, and the permissions of the files and directories will be shown as nobody:nobody, which is the default anonymous user3
• Therefore, if a customer enabled NFSv4.0 on an SVM and changed the client mount from NFSv3 to NFSv4, and found that the directory owner was changed from root to nobody, the most likely cause is that the ID mapping domains do not match between the client and server. The customer should check and correct the /etc/idmapd.conf file on both systems, and restart the idmapd service and remount the NFSv4 share4

References:

1: ONTAP 9 - Network File System (NFS) - The Open Group 2: ONTAP 9 - NFSv4 and NFSv4.1 Enhancements - The Open Group 3: NFSv4 mount incorrectly shows all files with ownership as nobody:nobody - Red Hat Customer Portal 4: NFSv4 mountpoint shows incorrect ownerships as nobody:nobody in CentOS/RHEL - The Geek Diary

 To assign storage access to an NVMe protocol host, you need to configure the following objects:

• The NVMe subsystem, which is a logical entity that represents a group of NVMe namespaces that can be accessed by one or more hosts1.
• An NVMe LIF, which is a logical interface that provides network connectivity for the NVMe subsystem2.
• The NVMe namespace, which is a logical volume that is exposed to the host as an NVMe device1.

You do not need to configure the NVMe broadcast domain or the NVMe LUN, as these are not relevant for NVMe protocol host access. References = 1: Configuring NVMe-oF - Hitachi Vantara Knowledge 2: Basic VMware NVMe Architecture and Components

FabricPool is an ONTAP feature that enables tiering of cold data from SSD aggregates to low-cost object storage, either on-premises or in the cloud1. FabricPool requires a license to be installed on the cluster, and the license type depends on the cloud tier being used2. In this scenario, the cloud tier is another ONTAP cluster (FAS8300), which is not supported by the new Cloud Tiering license that is used for most FabricPool configurations3. Therefore, the old FabricPool license that is retained for dark sites or MetroCluster systems using FabricPool Mirror must be used3. The FabricPool license defines the amount of capacity that can be tiered to the cloud tier, and it can be increased by add-on orders4. References:

• 1: FabricPool overview5
• 2: FabricPool requirements6
• 3: Install a FabricPool license2
• 4: ONTAP FabricPool (FP) Licensing Overview1

• IQN stands for iSCSI Qualified Name, which is a unique identifier for an iSCSI initiator or target1.
• ONTAP uses IQN to authenticate and authorize iSCSI sessions2.
• If two hosts have the same IQN, they will cause a conflict and ONTAP will reject the new session request from the second host3.
• This will result in the host losing the connection to the iSCSI target and then reporting that the session was reestablished, as shown in the exhibit.
• To avoid this problem, each host should have a unique IQN. References:
• iSCSI Qualified Name (IQN) - NetApp
• iSCSI authentication and authorization - NetApp
• Troubleshooting iSCSI issues - NetApp
• [Configuring iSCSI initiators - NetApp]

• The error message “wafl.cp.toolong:error” indicates that a WAFL consistency point (CP) took longer than 30 seconds to complete. A CP is a process that flushes the data from the NVRAM buffer to the disk. A long CP can cause latency and performance issues for the system1
• One possible cause for a long CP is excessive SSD load causing the wear leveling to become unbalanced. Wear leveling is a technique that distributes the write operations evenly across the SSD cells to extend the lifespan of the SSD. If some SSD cells are written more frequently than others, the wear leveling will become unbalanced and the SSD performance will degrade2
• Another possible cause for a long CP is an SSD disk performing garbage collection to create a dense data layout. Garbage collection is a process that reclaims the space occupied by invalid or deleted data on the SSD. Garbage collection can improve the write performance and storage efficiency of the SSD, but it can also consume CPU and disk resources and cause long CPs3
• A disk failing or being failed is not a likely cause for a long CP, because the system will automatically mark the disk as failed and remove it from the aggregate. The system will also initiate a disk reconstruction or a RAID scrub to restore the data protection and redundancy4
• There is no evidence that the system has SATA HDDs, so there is no reason to assume that there is excessive SATA HDD load. Moreover, SATA HDDs are usually used for secondary or backup storage, not for primary or performance-sensitive workloads5

References:

1: Are long Consistency Points (wafl.cp.toolong) normal? - NetApp Knowledge Base 2: How to troubleshoot SSD performance issues - NetApp Knowledge Base 3: How to troubleshoot SSD garbage collection issues - NetApp Knowledge Base 4: How to troubleshoot disk failures and replacements - NetApp Knowledge Base 5: ONTAP 9 - Hardware Universe - The Open Group