Short definitions of the IBM Power, IBM i, VMware, SAN and storage terms that XorMon monitors. Each entry names the manufacturer documentation it is based on and links to the matching XorMon page.
A logical partition (LPAR) lets one physical IBM Power server run as several independent servers. Processors, memory and I/O devices are split into subsets, and each subset runs its own operating system: AIX, IBM i, Linux or Virtual I/O Server.
Full definition · XorMon: IBM Power CPU monitoring · Source: IBM Docs, Logical partition overview (Power11)
A Hardware Management Console (HMC) is an appliance used to configure and control IBM Power servers and their partitions. If it fails, the server keeps running but its partitions cannot be reconfigured.
Full definition · XorMon: IBM Power monitoring · Source: IBM Docs, Hardware Management Console (Power10)
The Virtual I/O Server (VIOS) is software, part of PowerVM, that shares the physical I/O resources of a server among client partitions. It runs in its own partition and provides virtual SCSI, virtual Fibre Channel (NPIV) and Shared Ethernet Adapter services.
Full definition · XorMon: IBM Power monitoring · Source: IBM Docs, Virtual I/O Server overview (Power10)
NPIV lets several partitions share one physical Fibre Channel port, each reaching its own SAN storage through unique virtual WWPNs. The Virtual I/O Server links the virtual and physical adapters without accessing the storage itself.
XorMon: IBM Power SAN topology · Source: IBM Docs, Virtual Fibre Channel (Power10)
SR-IOV extends PCIe so that several partitions can share one adapter through virtual functions. A partition reaches its virtual function directly, without going through the hypervisor or a Virtual I/O Server, which lowers latency and CPU load.
XorMon: IBM Power LAN monitoring · Source: IBM Docs, Single root I/O virtualization (Power10)
IBM i is an IBM operating system that integrates database, middleware, security, runtime and hypervisor in a single product, licensed as one unit. XorMon lists IBM i systems as IBM i (AS400).
Full definition · XorMon: IBM i monitoring · Source: IBM, IBM i product page
A disk pool, also called an auxiliary storage pool (ASP) in the character-based interface of IBM i, is a software definition of a group of disk units that does not have to match the physical disk layout. Recovery after a disk failure is limited to the pool that held the failed unit.
XorMon: IBM i monitoring · Source: IBM Docs, Disk pools (IBM i 7.4)
On IBM i, a memory pool is a logical portion of main memory reserved for processing a job or a group of jobs. Each pool is private or shared, and up to 64 pools can be active at the same time.
XorMon: IBM i monitoring · Source: IBM Docs, Work management memory pool concepts (IBM i 7.6)
CPU Ready is the time the virtual CPUs of a VMware virtual machine wait for a physical CPU thread to be free. The vSphere Client shows it in milliseconds per interval, esxtop as a percentage (%RDY).
Full definition · XorMon: VMware CPU monitoring · Source: Broadcom, Understanding the CPU ready values in the vSphere Client advanced performance charts
CPU Co-Stop applies to virtual machines with more than one vCPU. It is the percentage of time a VM that was ready to run was held back by co-descheduling, shown as %CSTP in esxtop.
Full definition · XorMon: VMware CPU monitoring · Source: Broadcom, ESXTOP overview for Performance Troubleshooting
vSAN is a software layer built into the ESXi hypervisor. It pools the local or directly attached storage devices of the hosts in a cluster into one shared pool, so that HA, vMotion and DRS work without an external shared storage array.
Full definition · XorMon: VMware vSAN monitoring · Source: Broadcom, What Is vSAN (vSAN 8.0)
In vSphere, a datastore is a logical container, similar to a file system, that hides the details of the physical storage and holds virtual machine files. Supported types are VMFS, NFS, vSAN and Virtual Volumes.
Full definition · XorMon: VMware monitoring · Source: Broadcom, Working with Datastores in vSphere Storage Environment (vSphere 8.0)
AHV is the native hypervisor of Nutanix and the default option for Nutanix HCI. It combines virtualization, networking, infrastructure and operations management in a single interface, Nutanix Prism.
XorMon: Nutanix monitoring · Source: Nutanix, AHV Administration Guide (AHV 6.8)
Buffer-to-buffer (BB) credits are the flow-control mechanism between two directly connected Fibre Channel ports. A sender spends credits as it transmits and waits for the neighbor port to give them back.
Full definition · XorMon: Buffer-to-buffer credit monitoring · Source: Broadcom, Buffer-to-Buffer Flow Control (Fabric OS 9.2.x)
A SAN fabric is the part of the storage network made of switches and routers. It is divided into zones, and a device only communicates with the devices of its own zone.
XorMon: SAN monitoring · Source: IBM Docs, SAN fabric (FlashSystem 7300 and 7200 8.6.0)
An inter-switch link (ISL) is a direct link between two Fibre Channel switches, from E_Port to E_Port. To merge two fabrics, parameters such as R_A_TOV and E_D_TOV must be identical on both sides, and each switch needs its own domain ID.
A WWPN (worldwide port name) is a 16-digit hexadecimal name assigned to a port or to another SAN service. The equivalent name assigned to a node is the WWNN.
Source: IBM Docs, Worldwide node and port names (FlashSystem 5200 and 5000 8.6.0)
The optical power of an SFP transceiver is reported in dBm and microwatts. On Brocade switches, received (Rx) power is shown for smart SFPs and transmitted (Tx) power only for long-wave modules.
Full definition · XorMon: SFP power monitoring · Source: Broadcom, sfpShow (Fabric OS 9.2.x command reference)
IOPS counts the SCSI commands or NAS operations that a storage system handles in one second. The same amount of data gives a different number of IOPS depending on the block size: a 1 MB file read in 4 kB blocks is 256 read operations.
XorMon: Storage performance monitoring · Source: NetApp, What are IOPs?
Storage latency is the time between a request or command being issued and its response, measured in seconds, milliseconds or microseconds.
XorMon: Storage performance monitoring · Source: NetApp, What are common performance terms?
Throughput is the amount of data moved per unit of time over a channel, expressed in operations or in data volume per second. It is often confused with bandwidth, which is the theoretical maximum of the channel.
XorMon: Storage performance monitoring · Source: NetApp, What are common performance terms?
In IBM Storage Virtualize systems such as FlashSystem, a storage pool is a group of managed disks from which the capacity of volumes is allocated, in extents of a size fixed when the pool is created. A volume takes its capacity from a single pool.
XorMon: IBM FlashSystem capacity · Source: IBM Docs, Creating storage pools using the CLI (FlashSystem 5200 and 5000 8.6.0)
Thin provisioning lets you create logical volumes larger than the installed physical capacity. Physical space is allocated only when data is written.
XorMon: IBM FlashSystem capacity · Source: IBM Storage Modeller, Defining capacity
On IBM FlashSystem, data reduction pools shrink the stored data by compression, deduplication and thin provisioning, and can combine several methods on one volume. Space that is no longer used by the host is reclaimed with the SCSI unmap command.
XorMon: IBM FlashSystem capacity · Source: IBM Docs, Moving volumes to a data reduction pool (FlashSystem C200 9.1.0)
Policy-based replication is an asynchronous replication method between two IBM FlashSystem systems, driven by volume groups and replication policies. Any volume added to a volume group with a policy is replicated, and the system alerts when the recovery point objective is exceeded.
XorMon: IBM FlashSystem policy-based replication · Source: IBM Docs, Replication (FlashSystem 9500, 9200 and 9100 8.7.0)
A snapshot is a consistent point-in-time copy of a volume group that hosts cannot modify. On IBM FlashSystem it uses space in a storage pool, by default the pool of the parent volume.
XorMon: Storage snapshots · Source: IBM Docs, Snapshot properties (FlashSystem C200 9.1.0)
Capacity planning predicts the resources a workload needs to meet its performance targets: CPU, memory, storage and network bandwidth. It aims to avoid under-provisioning, which creates bottlenecks, and over-provisioning, which wastes cost.
Full definition · XorMon: Capacity monitoring · Source: Microsoft Learn, Architecture strategies for capacity planning (Azure Well-Architected Framework)