Project

Profile

Help

Issue #8996

closed

Mitigate cache invalidation issues with a reduced default expiration time.

Added by dalley almost 3 years ago. Updated over 2 years ago.

Status:
CLOSED - CURRENTRELEASE
Priority:
Normal
Assignee:
Category:
-
Sprint/Milestone:
Start date:
Due date:
Estimated time:
Severity:
2. Medium
Version:
Platform Release:
OS:
Triaged:
Yes
Groomed:
No
Sprint Candidate:
No
Tags:
Katello, Performance
Sprint:
Sprint 100
Quarter:

Description

If Redis goes down, Pulp can no longer invalidate / clear out the cache appropriately while it is offline, and when it comes back up it will be out of date.

We should mitigate this by reducing the default expiration time while investigating more comprehensive solutions.


Related issues

Copied to Pulp - Backport #9063: Backport 8996 "Reduce cache entry expiration TTL" to 3.14.zCLOSED - CURRENTRELEASEdaviddavis

Actions
Copied to Pulp - Story #9079: Investigate additional mitigations to cache invalidation issuesCLOSED - DUPLICATE

Actions
Actions #1

Updated by dalley almost 3 years ago

  • Priority changed from Normal to High
Actions #2

Updated by dalley almost 3 years ago

  • Tags Easy Fix added
Actions #3

Updated by dalley almost 3 years ago

  • Category set to Installer - Moved to GitHub issues
Actions #4

Updated by dalley almost 3 years ago

  • Severity changed from 2. Medium to 3. High
  • Sprint set to Sprint 100
Actions #5

Updated by dalley almost 3 years ago

  • Description updated (diff)
Actions #6

Updated by dalley almost 3 years ago

  • Subject changed from Disable Redis persistence to [EPIC] Cache invalidation issues
  • Category deleted (Installer - Moved to GitHub issues)
  • Priority changed from High to Normal
  • Severity changed from 3. High to 2. Medium
  • Sprint deleted (Sprint 100)
  • Tags Performance added
  • Tags deleted (Easy Fix)
Actions #7

Updated by dalley almost 3 years ago

  • Subject changed from [EPIC] Cache invalidation issues to Mitigate cache invalidation issues with a reduced default expiration time.
  • Description updated (diff)
Actions #8

Updated by pulpbot almost 3 years ago

  • Status changed from NEW to POST
Actions #9

Updated by dalley almost 3 years ago

  • Assignee set to dalley
Actions #10

Updated by dalley almost 3 years ago

  • Sprint/Milestone set to 3.14.2
Actions #11

Updated by dalley almost 3 years ago

  • Copied to Backport #9063: Backport 8996 "Reduce cache entry expiration TTL" to 3.14.z added
Actions #12

Updated by dkliban@redhat.com almost 3 years ago

  • Triaged changed from No to Yes
  • Sprint set to Sprint 100
Actions #13

Updated by daviddavis almost 3 years ago

  • Sprint/Milestone deleted (3.14.2)
Actions #14

Updated by dalley almost 3 years ago

  • Sprint/Milestone set to 3.14.2
Actions #15

Updated by ipanova@redhat.com almost 3 years ago

  • Sprint/Milestone changed from 3.14.2 to 3.15.0
Actions #16

Updated by dalley almost 3 years ago

  • Copied to Story #9079: Investigate additional mitigations to cache invalidation issues added

Added by dalley almost 3 years ago

Revision 017b0b1e | View on GitHub

Reduce the cache entry expiration TTL

If Redis goes down then its persisted data can get out of sync with Pulp, because cache invalidation events cannot occur. We should mitigate the impact by letting cache entries expire after a shorter period of time.

closes: #8996 https://pulp.plan.io/issues/8996

Actions #17

Updated by dalley almost 3 years ago

  • Status changed from POST to MODIFIED
Actions #18

Updated by pulpbot over 2 years ago

  • Status changed from MODIFIED to CLOSED - CURRENTRELEASE

Also available in: Atom PDF