• Icon: Epic Epic
    • Resolution: Unresolved
    • Icon: Major Major
    • None
    • None
    • None
    • None
    • Collect etcd performance
    • False
    • Hide

      None

      Show
      None
    • False
    • To Do
    • 67% To Do, 0% In Progress, 33% Done
    • 8

      USER STORY:

      • As a partner, I want to measure the etcd performance in my cluster so that I can make sure my customers will have appropriated infra resources allocated to control plane as recommended by Red Hat.
      • As a Red Hat VCSP reviewer, I want to ensure the partner running OCP in their provider is following the deploying using recommended infrastructure resources to control planes, so the reviewers (Cert-Ops, engineers) can have more confidence when reviewing the results applied by a partner to validate OCP to VCSP program.

      DESCRIPTION:

      Currently the etcd measurement is manual as described in this article: https://access.redhat.com/articles/7010406#effects-of-network-latency--jitter-on-etcd-4 

      The collector plugin (PID 99) can be used to collect this data, preferrebaly before the must-gather.

      GOAL:

      • OPCT-247: Measure and collect "Effects of Network latency & jitter on etcd" using ipperf with a test of minimum of 6 minutes between the control plane nodes. Example: run server in master-1, and run two tests using master-2 and master-3 as clients
      • OPCT-245: Measure and collect "Effects of Disk latency on etcd" with fio tests with more than one samples (maybe collect across all nodes?). For example 6 samples with interval of 10 seconds
      • OPCT-247: Measure and collect "Effects of consensus latency on etcd" watching the etcd information within two minutes with interval of N (10?) seconds
      • OPCT-246: Collect prometheus metrics for the period of the Conformance tests execution
      • Collect and save the timeline (start and end) for each test, it can be helpful to track issues in the logs and isolate possible problems in etcd pod logs when parsing it.
      • Create a summary of the data collected
      • Document how to explore the data/results

      ACCEPTANCE CRITERIA

      • Collect all the data required to cover the article [1]
      • Create documentation to help users when reviewing data collected by OPCT

      ENGINEERING REFERENCES:

      •  

       

            rhn-support-mrbraga Marco Braga
            rhn-support-mrbraga Marco Braga
            Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

              Created:
              Updated: