Viewing the cluster overview

In the left navigation pane of AOC, click the name of the managed ACOS cluster to go to the cluster Overview page, where you can view information about the cluster and the usage of all resources.

• Alert

  Displays the type and number of unresolved alerts within the cluster, as well as information about recently triggered alerts.

  Based on severity and urgency, alerts are divided into three types: Critical, Notice, and Info.

  To view details of each alert, click the widget to navigate to the Alert page. You can also click a specific alert to open its details panel. For instructions on alert rule settings and notification management, refer to Managing alerting features.

• Active-active status

  For an ACOS active-active cluster, this section will display its topology status. However, for an ACOS cluster without the active-active feature enabled, this widget will not be displayed.

  • The components and resource information of the active-active cluster, such as the number of hosts and virtual machines within each availability zone (using AVE virtualization).
  • The health status of the active-active cluster: Displays the health status of the components in the active-active cluster. Any alerts are displayed in this widget as well.
  • The operation information of the active-active components: You can click a component to view its operation information on the opened overview page.

• Network

  When an ACOS (AVE) cluster provides storage services to external compute nodes, this widget will display the access virtual IP configured for the cluster. Otherwise, this widget will not be displayed. Click Edit to modify the access virtual IP of the cluster.

• Storage efficiency

  Displays metrics related to cluster space reduction, including overall storage efficiency, effective capacity ratio, data reduction ratio, data reduction savings, and storage overprovisioning ratio.

  The overall storage efficiency of a cluster measures the effectiveness of storage capacity optimization achieved through erasure coding, snapshots, and thin provisioning. Overall storage efficiency refers to the ratio of the maximum capacity available for all data using replication to the storage capacity used by the cluster.

  Effective capacity ratio refers to the percentage of the total logical used capacity of all volumes in the cluster against the used storage capacity of the cluster. In other words, the effective capacity ratio is the ratio of the total logical capacity used by all volumes to the total storage capacity used by the cluster.

  The space reduction ratio refers to the percentage of space saved within the cluster by using erasure coding compared to using replication with the same redundancy level. Data reduction ratio is the ratio of the capacity used by replication for all data to the capacity (S) used by current storage redundancy policy (replication or erasure coding) for all data.

  Data reduction savings refers to the storage space saved in the cluster by using erasure coding compared to using replication with the same redundancy level. Data reduction savings are calculated as the capacity occupied by all data minus the capacity (S) used by the current storage redundancy policy (replication or erasure coding) for all data.

  Storage overprovisioning ratio refers to the ratio of the total storage space allocated to volumes to the total storage capacity of the cluster. Storage overprovisioning ratio is the ratio of the total storage space (Q) allocated to virtual volumes to the total storage capacity of the cluster, where the total storage space allocated to virtual volumes is related to the current storage redundancy policy (replication or erasure coding) of the virtual volumes.

  Note:

  For virtual volumes using erasure coding, if the configuration is K + M:

  • When M = 1, the replication policy with the same redundancy level is 2 replicas; when 1 < M ≤ 4, the replication policy with the same redundancy level is 3 replicas.
  • Assuming that the physical capacity used by the volume is S, the capacity required by the replication policy with the same redundancy level is S / (K + M) × K × number of replicas.

  For virtual machines using a replication policy, if the storage space allocated to the virtual volumes is Q, then Q = Logical capacity × replication factor.

• CPU

  Displays information about the current CPU resources of the cluster.

  • The upper-right corner of this widget shows:
    • Total frequency: The sum of the CPU frequencies of all hosts in the cluster, for example, 998.4 GHz.
    • Total cores: The sum of the physical cores of all hosts in the cluster, for example, 384 cores.
  • Click the drop-down symbol to select and view the following CPU information:
    • Usage: The CPU usage of the cluster.
    • CPU allocation: Displays the number of CPUs reserved for system services, the number of exclusive CPUs, and the number of CPUs available for sharing among all virtual machines with CPU exclusive disabled.
    • Shared allocation for active VMs: Displays two metrics, Active VM allocation and Shared CPU allocation.

      • Active VM allocation is calculated as the shared CPU allocation divided by the total number of CPUs across all hosts minus the number of exclusive CPUs and the number of CPUs reserved for system services, where Shared CPU allocation for active VMs refers to the total number of non-exclusive vCPUs assigned to all running and suspended virtual machines across all hosts in the cluster.

      • Shared CPU allocation is calculated as the number of CPUs available for sharing among all virtual machines with CPU exclusive disabled divided by the total number of CPUs across all hosts minus the number of exclusive CPUs and the number CPUs reserved for system services.

• Memory:

  Displays the memory usage of the cluster, including total memory, available memory, memory allocated to active virtual machines, and system reserved memory.

  • Total memory: Displays the total memory available to the operating systems of all hosts in the cluster.
  • Available memory: Displays the amount of memory remaining in the cluster that can be allocated to the active virtual machines (including running and suspended virtual machines) after excluding the memory reserved for system services and active virtual machines.
  • Provisioned and used memory: Displays the total amount of non-exclusive memory allocated to all running and suspended virtual machines on the hosts.
  • System reserved memory: Displays the amount of memory reserved for the system to run necessary services.

• Storage:

  Displays the storage capacity of the cluster, including total storage capacity, used capacity, invalid capacity, available capacity, and estimated maximum capacity allocation.

  Invalid space refers to the total data storage capacity provided by unhealthy but online physical disks, nodes with storage failures, and normally unmounted physical disks.

  The estimated available storage space depends on the selected storage policy. By default, the allocatable storage space is displayed when the cluster default redundancy policy is used. When the cluster supports erasure coding, you can select erasure coding and calculate space allocation based on the corresponding space utilization.

  • By replication: The logical space that can be allocated is the available space divided by the replication factor. For example, when the replication factor is two, the maximum storage space that can be allocated is 1/2 of the available space.
  • By erasure coding: Available allocatable space = Cluster available storage space × Space utilization corresponding to the erasure coding configuration. When the erasure coding configuration is K + M, the space usage is K / (K + M).

• Cache

  If the cluster uses tiered storage, the total cache capacity of the cluster, as well as the used, invalid, and available capacity, will be displayed.

  • Total cache:

    Used refers to the amount of cache partition capacity in use and its proportion relative to the total cache partition capacity.

    Invalid refers to the cache partition capacity that cannot be allocated, including the sum of the data space capacity provided by unhealthy and online physical disks, nodes with storage failures, and normally unmounted physical disks.

    Available refers to the remaining cache partition capacity that can be allocated. Available capacity is calculated as the total capacity minus the used capacity minus the invalid capacity.

  If the cluster is configured with a hybrid-flash storage setup or an all-flash setup with various types of SSDs, this widget will also display the read cache capacity and write cache capacity of all physical disks in the cluster. The capacities of both cache types are further categorized into Used, Invalid, and Available. The ratio of total read cache capacity to total write cache capacity is 2:8.

  • Read cache:

    Used refers to the amount of read cache partition capacity in use and its proportion relative to the total read cache partition capacity.

    Invalid refers to the read cache partition capacity that cannot be allocated.

    Available refers to the remaining read cache partition capacity that can be allocated. Available read cache capacity is calculated as the total read cache capacity minus the used read cache capacity minus the invalid read cache capacity.

  • Write cache:

    Used refers to the amount of write cache partition capacity in use and its proportion relative to the total write cache partition capacity.

    Invalid refers to the write cache partition capacity that cannot be allocated.

    Available refers to the remaining write cache partition capacity that can be allocated. Available write cache capacity is calculated as the total write cache capacity minus the used write cache capacity minus the invalid write cache capacity.

  If the cluster has volume pinning enabled, the total capacity reserved for pinned volumes and its percentage of the total cache capacity will also be displayed.

  • Volume pinning:

    Displays the total cache partition capacity reserved for pinned volumes, the cache partition capacity used by pinned volumes, the available cache partition capacity for pinned volumes, and the percentage of total capacity that each represents.

    Used refers to the total cache partition capacity used by pinned volumes. In some cases (such as disk removal), the used capacity may exceed the total reserved capacity. This may trigger an alert indicating that the pinned volumes in the cluster consuming excessive cache space.

    Available refers to the remaining cache partition capacity that can be allocated to pinned volumes. Available capacity (pinned volumes) is calculated as the total reserved capacity minus the used reserved capacity for pinned volumes. When the used capacity exceeds the total reserved capacity, the available capacity is 0.

    Allocatable capacity refers to the maximum logical capacity that can be allocated to pinned volumes calculated based on the specified replication factor.

• Storage performance

  Displays the storage performance of the cluster in read, write, and total (read + write) scenarios in the form of a line chart, with the following metrics included:

  • IOPS
  • I/O average latency
  • I/O bandwidth
  • I/O average block size

• Dynamic resource scheduler (DRS)

  Displays the resource load status of an ACOS (AVE) cluster when the DRS feature is enabled for that cluster.

  The DRS score of a cluster is the average DRS score of all running virtual machines in the cluster and is used to evaluate the overall resource balance status of the cluster. A higher DRS score indicates more balanced resource usage in the cluster.

  If there are currently unexecuted migration recommendations, you will also see the number of migration recommendations to be executed.

• Data recovery

  Displays the size of data undergoing data recovery in the cluster, as well as the recovery rate.

• VM

  Shows the total number of virtual machines in the ACOS (AVE) cluster, as well as the number of virtual machines in various states.

• Host

  Shows the total number of hosts in the cluster and the number of hosts in various states.

• Physical disk

  Shows the total number of HDD and SSD drives in the cluster and the number of drives in various states.

• Cluster

  Shows the version information, license type, software, CPU architecture, mode (RDMA/Boost), and other related information about the cluster.

• Recycle bin

  Shows the number of virtual machines in the recycle bin belonging to the current ACOS (AVE) cluster, the storage capacity they occupy, and the amount of space to be released today and in the next seven days.