Skip to main content
NetApp Knowledge Base

What is "PEAK_PERFORMANCE" line in Active IQ or the headroom in ONTAP?

  1. Last updated
  2. Save as PDF
Views:
1,934
Visibility:
Public
Votes:
1
Category:
ontap-9
Specialty:
perf
Last Updated:

Applies to

  • ONTAP 9
  • Active IQ (AIQUM)
  • CPU/DISK utilization

Answer

  • The peak performance line represents the limit after which performance may become degraded.
    • It is a calculated value based on a complex internal algorithm of your platform type and historical workload.
  • After resource ( CPU/DISK ) utilization reaches the peak performance line, the latency may increase. 
  • The difference between the current utilization and peak performance line represents performance capacity (HEADROOM).
    • Note: Active IQ Unified Manager's performance capacity graphs allow for more refined computation which can be done separately from ONTAP but have the same concept as ONTAP's onboard headroom calculations
  • To understand if the issue is based on CPU/DISK utilization see the following  KB: How do I know if CPU is causing a performance issue?

Additional Information

  • The resource_headroom_aggr and resource_headroom_cpu objects can be used to calculate headroom

Example: The aggregate headroom output is shown below

Note: The output is the same for CPU headroom

Cluster::> set -privilege diag

Warning: These diagnostic commands are for use by NetApp personnel only.
Do you want to continue? {y|n}: y

Cluster::*> statistics start -object resource_headroom_ (Tab key pushed here)
    resource_headroom_aggr resource_headroom_cpu
Cluster::*> statistics start -object resource_headroom_aggr
Statistics collection is being started for sample-id: sample_219
Cluster::*> statistics show

Object: resource_headroom_aggr
Instance: DISK_HDD_node1_1_FC_1_9efd97e7-9c33-4252-8598-88424de957cc
Start-time: 1/10/2023 23:09:53
End-time: 1/10/2023 23:09:56
Elapsed-time: 3s
Scope: node1

    Counter                                                     Value
    -------------------------------- --------------------------------
    current_driver_qtime                                          0us
    current_latency                                               0us
    current_ops                                                     0
    current_utilization                                            0%
    ewm_std_dev_daily                                               -
                                 ops                                9
                   optimal_point_ops                               37
                             latency                            35580
               optimal_point_latency                           374170
                         utilization                                9
           optimal_point_utilization                               26
    ewm_std_dev_hourly                                              -
                                 ops                                4
                   optimal_point_ops                               36
                             latency                            19376
               optimal_point_latency                           160386
                         utilization                                1
           optimal_point_utilization                               20
...
 

  • Optimal outputs are provided in groupings like 'ewm_std_dev_daily' and 'ewm_std_dev_hourly'.
  • The delta between the current and optimal values is the Headroom Available for the resource.
  • If the current values (or current_utilization in AIQUM) exceed the optimal point values (or peak_performance in AIQUM) it is expected to see potentially high increases in resource latency which could translate to increases in protocol/client data latency.
current_latency The current operation latency of the resource.  Units are microseconds.
current_ops The current throughput of the resource in ops.  This is derived from an approximation of client ops.  Units are ops/s. 
current_utilization The current utilization of the resource. Units are percent.
optimal_point_ops The optimal point of ops derived from the latency/utilzation curve for the given interval (daily, hourly, etc).  This is derived from an approximation of client ops. Units are ops/s. 
optimal_point_latency The optimal latency point of the latency/utilization curve for the given interval (daily, hourly, etc).  This is derived from network and system latency per volume workload.  Units are microseconds.
optimal_point_utilization The optimal utilization point of the latency/utilization curve for the given interval (daily, hourly, etc). Units are percent.  

 

 

NetApp provides no representations or warranties regarding the accuracy or reliability or serviceability of any information or recommendations provided in this publication or with respect to any results that may be obtained by the use of the information or observance of any recommendations provided herein. The information in this document is distributed AS IS and the use of this information or the implementation of any recommendations or techniques herein is a customer's responsibility and depends on the customer's ability to evaluate and integrate them into the customer's operational environment. This document and the information contained herein may be used solely in connection with the NetApp products discussed in this document.