Uploaded image for project: 'Red Hat Advanced Cluster Security'
  1. Red Hat Advanced Cluster Security
  2. ROX-30540

Remediate and Stabilize Hash Implementation

    • Icon: Epic Epic
    • Resolution: Unresolved
    • Icon: Major Major
    • 4.10.0
    • None
    • None
    • Remediate and Stabilize Hash Implementation
    • Quality / Stability / Reliability
    • L
    • False
    • Hide

      None

      Show
      None
    • False
    • Not Selected
    • In Progress
    • ROX-24278 - [ENG] Postgres Performance Enhancements and Search Framework Modernization
    • 20% To Do, 40% In Progress, 40% Done
    • Hide

      Progress:  We have updated what gets hashed to be inclusive instead of exclusive.  Which means we are not hashing as many things as we did before.  Relatively static things like Compliance Operator profiles we do not bother to dedupe.  We also do are not storing hashes for volatile things like node index reports that would essentially be different every time.  These changes have reduced the amount of things we hash.
      Additionally we have put in changes for things like with lists of items where the order is irrelevant.  Such objects were flagging as new when in essence they were actually dupes.  So this as well helps with the amount of data we are storing into the hashes table.

      Problem:

      Plan:  With the information provided above and the extra metrics that were added into 4.9, we recommend pushing this to 4.11 (or later) after we see what the real impact of the changes above are.

      Show
      Progress:  We have updated what gets hashed to be inclusive instead of exclusive.  Which means we are not hashing as many things as we did before.  Relatively static things like Compliance Operator profiles we do not bother to dedupe.  We also do are not storing hashes for volatile things like node index reports that would essentially be different every time.  These changes have reduced the amount of things we hash. Additionally we have put in changes for things like with lists of items where the order is irrelevant.  Such objects were flagging as new when in essence they were actually dupes.  So this as well helps with the amount of data we are storing into the hashes table. Problem: Plan:  With the information provided above and the extra metrics that were added into 4.9, we recommend pushing this to 4.11 (or later) after we see what the real impact of the changes above are.

      Fix design of the hashes table.

       

              rh-ee-dashrews David Shrewsberry
              rh-ee-dashrews David Shrewsberry
              ACS Core Workflows
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated: