Versions Compared

Old Version 2

changes.mady.by.user Scott Kirvan (Unlicensed)

Saved on

compared with

New Version Current

changes.mady.by.user Scott Kirvan (Unlicensed)

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

All the individual improvement ideas are tracked in Jira epic DAOS-3929.


Ticket(s)TaskStatusResult

DAOS-3914

CORCI-840

Increase throughput on HW clusters by running multiple servers per HW node and split one of the HW clusters.in-progressCORCI-841Build new HW clusters and get them into CI as soon as HW shows upblockedDAOS-3915Analyze HW tests to see if they can be moved to weekly and to understand if different cluster combos would helpDone

Splitting clusters from 8 to 4 node will no make an

immediate substantial impact.  This is going to help 

long term though, particularly when adding the new 

hardware.

CORCI-717Node failures are often the cause of intermittent CI issues but need console log to debug
todoCORCI-842PDT presentation on commit message pragmas, use of quickbuild
todo
DAOS-3868
Fix quickbuild issues, this will allow short-circuit of time consuming build operations when appropriatetodo

CORCI-711HW nodes are provisioned from snapshot, fix out-of-space and other cruft issues that cause intermittent failurestodo

DAOS-3607

DAOS-2759

Increase flexibility to run different groupings of tests by running from RPMsin-progress
CORCI-843If a change impacts files in the "doc" directory then skip unnecessary build/test cycles.in-progress
todoDAOS-3921Reduce wait time in daos_test rebuild subtestsin-progressDAOS-3930IorSmall runtimes are all over the maptodo

The primary metric for tracking improvements is total hours master jobs take to complete.  Total time includes waiting time plus run time.  The chart below includes randomly select successful runs against master.

...


DCO-9105Add a jira field for bugs that can be incremented everytime the issue is seentodo

Investigate GitHub Checks as an alternative to the commit statuses that we currently usetodo

Investigate moving more error reporting out of Jenkins Workflow steps and into JUnit resultstodo

Data-mine Jenkins for statistics on how many PR commits are test retries  as a metric of how often we are hitting intermittent failures

  • see if we can determine the failure
todo

Data-mine Jenkins for statistics on how many PR commits have actual patches to rectify failed tests

  • see if we can determine which tests are catching failures the most frequently and least frequently
todo