Skip to main content
NetApp Knowledgebase

Performance Archives: Performance first failure data collection for Clustered Data ONTAP 8.3 and later

Views:
1,044
Visibility:
Public
Votes:
2
Category:
data-ontap-8
Specialty:
perf
Last Updated:

Applies to

  • ONTAP Select 
  • Clustered Data ONTAP 8.3 

Answer

  • What is a Performance Archive?
    Beginning with clustered Data ONTAP 8.3, performance content is collected and stored ('archived') on every system by default. This content is built on top of a new functionality in Data ONTAP, which automatically stores Quality of Service performance statistics on a per-volume basis for diagnostic purposes for up to 28 days on each clustered Data ONTAP node.

    Using the AutoSupport(tm) infrastructure, a Performance Archive can be created and uploaded to NetApp.
     
  • How Do I verify if Performance Archive is enabled?
    ::> set diagnostic
    ::*> statistics archive config show

verify field:
Is the Performance Archive Enabled?: true

if false, enable it by running the following command
::*> statistics archive config modify -is-enabled true

Note: EMS events will also report the following message when the Performance Archive is disabled
::*> event log show  -event perf*
Time                Node             Severity      Event
------------------- ---------------- ------------- ---------------------------
INFORMATIONAL perf.ccma.off: Performance archiver is not enabled.

  • How do I upload a Performance Archive?
    Run the 'system node autosupport invoke-performance-archive' sub-command with the following options:
    • -start-date <"MM/DD/YYYY HH:MM:SS">
    • -duration <[ h][ m][ s]> (or -end-date)
    • -node *
    • -case-number
       

It is important to note that the maximum duration for any single collection is 6 hours and the recommended sample period is 4 hours.

Example:
Using an external monitoring tool (example: OnCommand Performance Manager), a change in the latency trend for a cluster was identified between 6:00 AM and noon on 1 November 2014. According to the tool, an increase in latency occurred at 8:00 AM and lasted approximately 4 hours. As the best data set covers the period before the issue is seen as well as the transition to when the latency trend changes, data should be collected from 6:00 AM to 10:00 AM.

This would provide a 2 hour sample of data before the issue was identified (6:00 AM to 8:00 AM) and a 2 hour period (8:00 AM to 10:00 AM) when the issue was occurring. On reaching out to Customer Support, this issue was assigned case number 2008123456:

system node autosupport invoke-performance-archive -start-date "11/01/2020 06:00:00" -duration 4h -case-number 2008123456 -node *

Note: Start date should be in quotes, and follow the "MM/DD/YYYY" format. For example, September 1, 2019 would be "09/01/2019". Case number must be included, and collection must go from all nodes.

  • Why might I still need to run Perfstat?
    In order to ensure that issues are resolved as expediently as possible, it may be necessary to continue to collect Perfstat as well as the clustered Data ONTAP Performance Archive content simultaneously.

    A Performance Archive is a data dense binary file that requires post-processing to properly consume the content. Certain counters may not be included in the Archive, in early versions of ONTAP 9, and are only available via the CLI or via perfstat.
     
  • How can a user look at the content within a Performance Archive?
    OnCommand Performance Manager utilizes the underlying data set that is stored within a Performance Archive. Refer Setting up a connection between a Performance Manager server and an external data provider
     
  • How can I validate that a Performance Archive has been uploaded properly?
    The command system autosupport history show will give the status of the Performance Archive AutoSupport. It should show sent-successful on completion of the transmission. Additionally, AutoSupport messages will appear on your Active IQ page. Search by your serial number to verify if the Performance Archive has completed successfully and been processed by NetApp's systems.
     
  • A Performance Archive failed to upload. How do I resolve the failure?
    See KB Troubleshooting Workflow: AutoSupport messages are not received by NetApp (HTTP/SMTP) or partner (SMTP)

    Note: Performance Archives can only be uploaded via HTTP/HTTPS.  SMTP is not supported as an upload method for Performance Archives. If you do not have HTTPS access to support.netapp.com for uploading archives, please see KB How to use OneCollect to upload performance archives to NetApp Support.

Additional Information

additionalInformation_text