Performance metadata for storage systems that run IBM Spectrum® Virtualize

View the lists of the performance metadata for IBM® SAN Volume Controller, IBM Spectrum Virtualize for Public Cloud, IBM Spectrum Virtualize as Software Only, IBM Storwize® storage systems, and for IBM FlashSystem® devices that run IBM Spectrum Virtualize.

Overview

The performance metadata is divided into the following categories:

Volume metadata

The following key metadata is collected for volume performance:
  • Data Rate (Read, Total, Unmap, Write)
  • Overall Host Attributed Response Time Percentage
  • Overall I/O Rate (Read, Total, Unmap, Write)
  • Pool Activity Score
  • Response Time (Overall, Read, Unmap, Write)
  • Volume Utilization
  • Write Cache Delay Percentage
The following I/O rate metadata is collected for volume performance:
  • Transfer Rate (Cache-to-Disk, Disk-to-Cache)
  • Unaligned Unmap I/O Rate
  • Write-Cache Delay I/O Rate
The following cache hit percentage metadata is collected for volume performance:
  • Overall I/O Cache Hits (Read, Total, Write)
The following response time metadata is collected for volume performance:
  • Peak Response Time (Read, Unmap, Write)
The following remote mirror metadata is collected for volume performance:
  • Global Mirror (Overlapping Write I/O Rate, Overlapping Write Percentage, Secondary Write Lag, Write I/O Rate)
The following volume cache metadata is collected for volume performance:
  • Cache Hits (Dirty Writes, Read, Total, Write)
  • I/O Rate (Destage, Read, Total, Write)
  • Response Time (Destage, Stage)
  • Transfer Rates (Cache-to-Disk, Disk-to-Cache)
  • Write Delay Percentage (Flush-through, Total Delay, Write-through)
  • Write Delay Rate (Flush-through, Total Delay, Write-through)
  • Data Rate (Fast-Write Writes)

The following volume copy cache metadata is collected for volume performance:
  • Cache Hits (Dirty Writes, Read, Read-ahead, Total, Write)
  • I/O Rate (Destage, Prestage, Stage, Total)
  • Response Time (Destage, Prestage, Stage)
  • Transfer Rates (Cache-to-Disk, Disk-to-Cache)
  • Write Delay Percentage (Flush-through, Total Delay, Write-through)
  • Write Delay Rate (Flush-through, Total Delay, Write-through)
  • Data Rate (Fast-Write Writes)

The following compression metadata is collected for volume performance:
  • Compressed Volumes (Data Rate, I/O Rate, Response Time)
  • Uncompressed Volumes (Data Rate, I/O Rate, Response Time)
The following miscellaneous metadata is collected for volume performance:
  • Cache to Host Transfer Response Time
  • Non-Preferred Node Usage Percentage
  • Transfer Size (Overall, Read, Write)
  • Unaligned Write I/O Rate
The following legacy cache metadata is collected for volume performance:
  • Dirty Write Percentage of Cache Hits
  • Write-Cache I/O Rate (Flush-through, Overflow, Write-through)
  • Write-Cache Percentage (Flush-through, Overflow, Write-through)

Disk metadata

The following key metadata is collected for disk performance:
  • Data Rate (Read, Total, Write)
  • I/O Rate (Read, Total, Write)
  • Response Time (Overall, Read, Write)
The following response time metadata is collected for disk performance:
  • Peak Back-end Response Time (Read, Write)
  • Peak Back-end Queue Time (Read, Write)
  • Queue Time (Overall, Read, Write)
The following miscellaneous metadata for disk performance:
  • Cache Destage (In-Flight I/O, Target I/O)
  • Transfer Size (Overall, Read, Write)

Pool metadata

The following key metadata is collected for pool performance:
  • Max Write Cache Fullness
  • Write Cache Fullness

Port metadata

The following key metadata is collected for port performance:
  • Data Rate (Receive, Send, Total)
  • I/O Rate (Receive, Send, Total)
The following I/O rate metadata is collected for port performance:
  • Port-to-Disk I/O Rate (Receive, Send, Total)
  • Port-to-Host I/O Rate (Receive, Send, Total)
  • Port-to-Local Node I/O Rate (Receive, Send, Total)
  • Port-to-Remote Node I/O Rate (Receive, Send, Total)
The following data rate metadata is collected for port performance:
  • Port-to-Disk Data Rate (Receive, Send, Total)
  • Port-to-Host Data Rate (Receive, Send, Total)
  • Port-to-Local Node Data Rate (Receive, Send, Total)
  • Port-to-Remote Node Data Rate (Receive, Send, Total)
The following response time metadata is collected for port performance:
  • Port-to-Local Node Response Time (Overall, Receive, Send)
  • Port-to-Remote Node Response Time (Overall, Receive, Send)
The following error rate metadata is collected for port performance:
  • Frame Errors (CRC Error Rate)
  • Link Errors (Invalid Link Transmission Rate, Invalid Transmission Word Rate, Link Failure Rate, Primitive Sequence Protocol Error Rate, Loss of Signal Rate, Loss of Sync Rate)
  • Port Congestion Index
  • Port Protocol Errors (Port Send Delay I/O Percentage, Port Send Delay Time, Zero Buffer Credit Percentage, Zero Buffer Credit Timer)
The following miscellaneous metadata is collected for port performance:
  • Port-to-Local Node Queue Time (Overall, Receive, Send)
  • Port-to-Remote Node Queue Time (Overall, Receive, Send)

Node metadata

The following metadata is collected for node performance:
  • Cache Fullness (Max Read Cache Fullness, Max Write Cache Fullness, Read Cache Fullness, Write Cache Fullness)
  • Compression CPU Utilization (Core 1 to Core 8)
  • CPU Utilization (Compression CPU, System CPU)
  • Garbage Collection (Data Movement Rate, Data Rewrite Rate, Extent Collection Rate, New Address Write Rate, Reclaimable Capacity, Recovered Capacity Rate)
  • Node Utilization by Node
  • System CPU Utilization (Core 1 to Core 8)